Adaptive mapping for improved pitch accuracy on touch user interfaces

Touch user interfaces such as touchpad or pen tablet are often used forcontinuous pitch control in synthesis devices. Usually, pitch is set at thecontact point on the interface, thus introducing possible pitch inaccuracies atthe note onset. This paper proposes a new algorithm, based on an adaptiveattraction mapping, for improving initial pitch accuracy with touch userinterfaces with continuous control. At each new contact on the interface, thealgorithm adjusts the mapping to produce the most likely targeted note of thescale in the vicinity of the contact point. Then, pitch remains continuouslyadjustable as long as the contact is maintained, allowing for vibrato,portamento and other subtle melodic control. The results of experimentscomparing the users' pitch accuracy with and without the help of the algorithmshow that such a correction enables to play sharply in tune at the contact withthe interface, regardless the musical background of the player. Therefore, thedynamic mapping algorithm allows for a clean and accurate attack when playing touch user interfaces for controlling continuous pitch instruments like voicesynthesizers.