Why Audio Testing is Making Smart Devices Lose Their Smarts, with Daniel Knighten of Audio Precision

When we hear the term “smart” device, words that may come into our minds are easy, convenient, or technologically advanced, but consumers and designers alike underestimate the challenges smart devices may create for the companies that are trying to test them, especially in the audio realm.

On today’s Pro AV podcast, we hear from Daniel Knighten, vice president of product development for Audio Precision. “Smart” from Knighten’s perspective has “translated into adding a speaker and microphone to every device under the sun,” including refrigerators and even toilets, Knighten said. He gets a lot of calls from companies who have never dealt with audio devices before who are now trying to test audio signals in devices that may not have specifically been built to support them.

Knighten explained that audio experts are used to dealing with devices where there’s an input and output and traditional audio measurement where testers can simply input an audio signal and check for the output of that signal to see how it’s changed. In a cloud-based atmosphere, companies like Apple, for example, have stated that they will never release audio that’s recorded by a HomePod; it’s become a real challenge to learn the figures of the input the way the machine “hears” it when it’s almost impossible to inject or extract a test signal.

These “audio roach motels,” as Knighten calls them, remove the simplicity from testing; testers have to upload a variety of noises into the cloud, get them onto a built in audio player, then order the smart device to play them in open-loop testing. It has become so inconvenient that Knighten thinks some of these tests may be suffering, particularly when the machine itself changes the variables by adapting to different environments automatically.