# Pitch Dithering:

As of late 2010 I have been using a technique I have devised to move from one kind of harmonic area, or one specific pitch centre, to another. This technique (which at present I call simply Pitch Dithering) has been facilitated in part by OpenMusic (OM). Allow me to use an analogy:

This diagram illustrates two colours side by side − in the first box there is a hard line − the content does not overlap at all. By the fourth box, the two contents are interlaced, and as such the lines become blurred.

The best way for me to achieve this at the moment is to use OM to do this for me. Imagine two types of content: A **B**

There are various ways to phase between the two; one can employ some kind of pattern, such as:

AAAAAAAAAA

A**B**AAAAAAAA

A**B**A**B**A**B**A**B**A**B**A

A**BB**A**B**A**B**A**B**A**B**

A**BB**A**BB**A**B**A**B**A

A**BB**A**BB**A**BB**A**B**

and so on, or, another way in which a similar effect can be achieved is by allowing a random process to oscillate between the two materials. Of these two techniques, my preference is to oscillate randomly between two materials, over a certain amount of time, but with controlled distribution. In other words, imagine I wish to get from A to B in 30 seconds, and I am in the middle of a dense hemi−demi−semiquaver passage. Furthermore, let me say that my crotchet pulse is at 70. At 70bmp, there are almost 35 crotchet beats, which is 182 hemi-demi-semiquavers. I want to go from material A (which is 100% distribution in favour of A) to material B (which will be 0% distribution in favour of A) − it is safe to assume that everything which goes immediately before the pitch dithering is A, and everything after B.

To return to our 182 hemi−demi−semiquavers, we want to move from 100% in favour of A, to 0%, over 30 seconds. For the distribution to remain roughly even, we can adjust the favour of A, from 0% to 100%, in ten steps (for convenience, it could be any number of steps really). So, the first distribution is 90% in favour of A: **i**: this generates a random number between 0 100;**ii**: this checks to see if the number generated by i is less than 90;**iii**: if it is less than 90, then select material at random from a, otherwise selected material at random from b;**iv**: repeat this procedure 18 times.

This is then repeated ten times, each time reducing the favour of A:

Here is an example, where a harmonic series (with microtonal invariance) is dithered against a whole-tone scale: