Listening tests allow us to find out how humans judge the results of an audio engineering process or product. In general we audio engineers are more interested in finding out something about the thing that’s being evaluated than we are about the human perceptual process, although we may be interested in that too. It makes us a bit different to hearing psychologists.

For this month’s Inside Track I’ve compiled a list of links to papers and other resources that might get you thinking about some of the fundamental challenges arising when conducting serious listening tests. By that I mean trying to work in a scientific way rather than just getting a few people together to listen to something and saying "yes that sounds nice" or "I prefer the one with the gold plated mains plugs." Some of the papers I’ve chosen are older ones, and some more recent, but they all try to tackle some aspect of the overall problem. What are we trying to evaluate and why? What biases might influence the results? How can things like learning and culture change the outcomes, and what happens if you can see the things you’re testing? How to select and (maybe) train listeners, and how to select programme material, these are both important factors. Then there’s the thorny question of how to differentiate between what people like and how they describe it, or what is "correct." Do people mean the same things by certain words used to describe sound, and how can we develop a common language for the purpose? Finally how might we map information about what people like onto information about how things sound?

I’m not claiming that these are the definitive papers on any of these topics, but they are AES convention, conference and journal papers that demonstrate and discuss some of the important points if you’re going to embark on a serious project of this kind. There are also videos, standards and other resources that might be useful.