"I do a fair amount of subjective testing in the audio field, in addition to my audio production and studio design work. Over the past six months, I’ve gotten caught up in a food fight on the Internet about the audibility of various audio processes.
That food fight really got me to thinking about the trouble we all have communicating about these issues. There are two central parts to the problem. First, we aren’t very objective about describing our subjective impressions and, second, we mostly fail to consider what is generally meant by the terms that we use. Such foibles are perfectly normal for us audio engineer types, and because we aren’t scientists by profession we can’t be held responsible for not making the kind of careful objective claims that real science requires.
Nevertheless, it is important for us to be able to understand at least a little bit about what is really going on in regard to the concept of audibility, just to get along in our professional audio lives. At the same time, we hear a lot of talk about the need for expanded resolution in digital audio these days, which raises some important questions. Can we really hear the difference? And, how important is it?
The answers we get to these questions when we carefully read the trade journals and the reported findings about audibility are a little confusing and contradictory. People have described the difference between 16-bit and 20-bit audio as the difference between mediocre and awesome. Others have ascribed a remarkable transparency to 20-bit signals. Similarly, individuals have reported increased clarity, definition, detailing and other virtues to audio signals with sample rates significantly higher than 48 kHz. Meanwhile, others note that such observations do not get made when we use double-blind tests, nor are they supported by such blind tests, presumed to be more rigorous and “objective” than more informal studies. In fact, double blind tests often seem to show that listeners can’t reliably distinguish between 16-bit audio and 20-bit audio.
How can this be? How can people clearly hear an effect, to a point where they choose to describe it with a superlative such as “amazing,” when under controlled blind conditions other people (and sometimes they themselves) can’t distinguish it from the same signal without said “amazing” effect? One cynical answer is that they are simply making it up. This is called the “Emperor’s New Clothes” syndrome. Humans are suggestible, and it is pretty easy for us to get stampeded into a group-think state of mind where we will clearly hear anything we think will be socially acceptable, whether it’s real or not. All of us working in audio have had the experience of equalizing a channel to a point where everybody in the control room agrees we’ve made some really significant, perhaps awesome, improvements, only to discover that we were equalizing an adjacent channel that wasn’t even switched on!
But I’m not satisfied with “The Emperor’s New Stereo” answer. I’m sure that suggestibility is often an issue, but we know very well that we are susceptible to making these mistakes and we’re generally pretty careful about avoiding them. Further, too many people I know whose hearing acuity I really respect have reported hearing things like a “BIIIIG” difference between 16-bit and 20-bit audio for me to say, “Nah, that’s just group-think. They’re making it up.”
One of my goals, as a researcher, is to find out how to resolve these differences, how to reconcile apparently contradictory and mutually exclusive reports about the audible effect of some audio process. I’d like to be able to explain how it can be that these differences exist, rather than assuming and trying to prove that one of the viewpoints is simply “wrong.”
There are several things to think about.
•First, we need to consider the actual nature of blind tests.
•Second we need to think about what we really mean when we describe an improvement as “amazing.”
•Third, we need to know how the term “audible” is really defined.
•Fourth, we need to know something about the nature of human perception.
Over the next several months, we’ll look into these questions in detail. There’s actually quite a bit to be said about each of the four questions I’ve raised. But first, we need to consider why this stuff is so important.
We are involved in a business of creating illusions for our customers. We spend a great deal of time and money on these illusions to make them effective. Our viewers/listeners similarly spend a great deal of money to experience illusions they find compelling, convincing and satisfying. At the same time, we’ve bought into the notion that the illusions become better as they become “more accurate,” and so we keep trying to improve the resolution of our illusions.
Meanwhile, it’s a competitive world out there. He/she who creates the most powerful illusion for the least bucks wins. We need to not waste our time and resources on effects that don't matter. We need to know what matters, how much it matters and how much it costs. Only then can we begin to really get into enhancing the illusion we are creating in a really meaningful way.
With that in mind, we’ll dive into the weird and wacky world of blind testing next month. Thanks for listening."
Link: Moulton Laboratories
0 comentarios:
Publicar un comentario en la entrada