A 2007 study by Sandra Pauletto and Andy Hunt at the University of York suggested that users were able to detect attributes such as noise, repetitive elements, regular oscillations, discontinuities, and signal power in audification of time-series data to a degree comparable with visual inspection of spectrograms.
[5] An example is the esophageal stethoscope, which amplifies naturally occurring sound without conveying inherently noiseless variables such as the result of gas analysis.
[13] Both sonification and audification are representational techniques in which data sets or its selected features are mapped into audio signals.
[14] However, audification is a kind of sonification, a term which encompasses all techniques for representing data in non-speech audio.
[citation needed] Their relationship can be demonstrated in the way data values in some sonifications that directly define audio signals are called audification.