Quantifying Touch
A data-oriented HCI research study exploring the analog vs. digital debate in the music space. This Master's project at UC Santa Cruz examined whether hardware or software input methods produce different accuracy outcomes.
Problem
There is a longstanding divide among musicians: those who believe digital audio solutions are optimal, and those who believe analog hardware predecessors are ideal. The arguments on both sides are largely subjective — hardware users cite tactile benefits and warmth, while software users point to computational power and adaptability. No one had put this to an objective, data-driven test.
Methodology
A 40-user study was devised comparing MIDI controller input versus mouse input. Participants completed two tests — amplitude accuracy and frequency accuracy — using a custom GUI built in SuperCollider. Each participant was randomly assigned an input method and underwent 10 trials per test. Behavioral metrics including button press frequency and trial timing were also recorded.
Key Findings
The input method (hardware vs. software) did NOT significantly affect user accuracy in either test. However, in the amplitude test, time spent on each trial correlated with better accuracy (p=.001). In the frequency test, the number of reference tone playbacks correlated with reduced error (p=.036). These secondary findings revealed interesting behavioral patterns even though the primary hypothesis was not supported.
Outcome
The research was completed as a Master's project at UC Santa Cruz, advised by Patrick Mantey, with collaboration from David Kant (UCSC Music) and Douglas Bonnett (Statistics). The full whitepaper is available for download.