In the audio community, you sometimes hear about people with a set of golden ears who can determine the frequencies and sound quality with machine-like precision. But here at Reviewed.com, we think it’s sub-optimal to hope for machine-like precision when you can have actual machine precision. It’s always a good idea to calibrate the scoring and criteria using expert opinions, but for the actual testing? We'd rather avoid opinions and instead have a system with accuracy and repeatability. So instead of our own ears, we use a special pair on "HATS."
The HATS (head-and-torso-simulator) is a analog, or stand-in, for the ear, used for data collection. We often use analogs of human senses in a many of our tests. For example a spectrophotometer—a human eye analog—is used to measure everything color and light related in our labs, from television screens to fabric stains. So for headphone testing, we use HATS to stand in not just for ears, but for the entire upper body. Let’s take a closer look at this analog.
More Than Just Ears
Intuitively, it might seem like a good idea to just use microphones to test headphones. But since human interactions with sound are so complex, evaluating a set of headphones solely by measuring sound traveling directly from the headphone to a microphone would yield severely unrealistic results. Human upper body architecture is quite complex, filled with tons of air-filled semi-enclosed spaces. All of these spaces affect the sound quality, for better or worse, by providing peculiar spaces for reverberation, all contributing to the total experience of perceiving a sound.
An intricate solution to a complex problem
Reverberation presents one of the largest problems when trying to collect objective data on a subjective experience such as sound. But since it’s such a complex phenomenon, it is extremely difficult—if not impossible—to model mathematically. This leaves us with one option: build an abstract human being with integrated microphones for eardrums. Lucky for us, we aren’t the first to be interested in accurate, reality-based sound quality determination. It turns out that acoustically abstract humans are available for purchase in the form of the Brüel & Kjær Head and Torso Simulator (HATS). They are used to measure auditory data in sound quality labs across countless industries, from car to washing machine manufacturers and, of course, the high-end headphone makers.
The HATS is a half mannequin that has the same properties as an average human, at least in terms of acoustics. Its torso is molded plastic in the shape of an average human torso and is lined on the inside to simulate a thoracic cavity, including separate lungs and a diaphragm. Connected via a short twistable neck atop the torso, a molded plastic head contains all acoustically relevant aspects of the human skull, such as the sinuses and mouth cavity. The mouth cavity contains a speaker, which can be used to produce sound when needing to simulate human speech. Attached to the head are two ear-shaped pieces of silicone that direct sound waves into anatomically correct ear canals, which terminate at high-quality microphones surrounded by molded silicone to simulate the human eardrum.
HATS in Action
This simulated human being has been a fixture in our headphone-testing lab since the beginning. We have placed countless headphones over, on or into HATS’ silicon ears and flooded its artificial ear canals with sound of every frequency audible to humans a thousand times over. These sounds reverberate in its head and lung cavities, most inevitably ending up collected in the microphones buried at the end of its ear canals after being slightly altered on their journeys. The data collected by the microphone is then analyzed and scored, giving you, the reader, the best possible idea as to the headphones’ performance outside of actually trying them for yourself.