In Tape Op #86, Greg Calbi related his early experiences as a mastering engineer. I was struck by his comment that professional mix engineers would ask him if a bass or vocal was too loud or too soft. Greg explained that what sounds right or wrong when making such decisions is "fluid." Chick Corea expressed a similar opinion in a recent Keyboard Magazine interview, stating that artistic judgment varies from one moment to the next. Everyone has experienced creating a mix that sounds stellar at the time, only to realize the next day that it actually sounds like crap. The frailty of auditory perception affects both amateurs and professionals alike.
Hearing fallibility has implications beyond doubting our own mixing decisions. It's the main reason audiophiles believe the sound of their hi-fis improves after replacing one perfectly competent AC power cord with another, or after raising their speaker wires off the floor on special "anti-resonance" cable lifters. Understand that I'm not talking about real differences, such as the sound of modern digital recording versus analog tape or vinyl records. Those differences are real, and are easily measured using the standard metrics of frequency response and distortion. Skilled listeners can easily identify which is which, nearly every time, in a blind test. Rather, what I'm addressing are differences that are perceived but not real — the placebo effect, if you will, or perhaps wishful thinking.
As much as we'd like to believe otherwise, our hearing memory is surprisingly short term. According to former DTS chief scientist James Johnston [1], hearing memory is accurate for about 1/4 of a second. He also explains that we can't focus on everything in a piece of music all at once — for example, on one playback we might notice the bass timbre, but ignore the quality of the ride cymbal. This makes it very difficult to know if subtle differences are real or imagined. If you play the same section of music five times in a row, the sound reaching your ears won't change unless you move your head, but the way you perceive each playback can and will vary. Is that delicate cymbal ping really clearer after replacing the power supply capacitors in your mixing console, or does it just seem that way?
Hearing limitations extend beyond our inability to pay attention to multiple details at once or our ability to remember the exact tonality of an instrument for longer than a few seconds. Another important hearing foible is the "masking effect," which conceals details in one source when similar frequencies are present in another. The most significant thing that happens when tracks are summed is psychoacoustic. An electric bass track where every note can be distinguished clearly when soloed might turn into rumbling mush after adding a chunky sounding rhythm guitar or piano. I'm convinced this is the real reason people wrongly accuse "summing" or "stacking" for a lack of clarity in their mixes. Masking occurs whether tracks are mixed with analog circuitry or an equivalent series of sample numbers are added in a DAW program or digital hardware mixer. The loss of clarity occurs in our ears and brain, not the summing device.
The hostility expressed in audio forums over these issues is discouraging. The fastest way to get a thread locked is to title it "converter loop-back test" and describe how nobody was able to hear the difference between a budget converter and one costing $1,000 per channel. A thread that asks "Do plug- ins sound as good as outboard gear?" is guaranteed to garner thousands of posts, many containing personal insults.
It doesn't have to be that way.
In my experience, the more someone knows about basic electronics — and proper audio test procedures — the less likely they are to believe that "science" hasn't found a way to measure what they're certain they can hear. Once you understand why the same wave file played from different hard drives cannot sound different (unless something is broken), you can move on to more important things that really do affect audio quality, such as mic placement and room acoustics. Even though some people insist there's more to audio fidelity than is currently known, this is easily disproved with a null test.
Null tests work by subtracting two audio signals against each other and analyzing what remains. If the result is total silence, then the signals are, by definition, identical. And if there is a residual, its audibility can be assessed based on its loudness and frequency distribution. The beauty of null tests is that they reveal all differences between two audio signals, including differences you might not even be looking for. So if there were some as-yet unknown property of audio fidelity, it would have been revealed years ago in a null test.
Listening tests can be valid, but they have to be done correctly. Again, I'm not talking about obvious differences that anyone can notice. And I'm definitely not talking about subjective preferences for which there is no right or wrong. One example of testing subtle differences is to assess whether recording at 96 kHz sounds clearer than at 44.1 kHz. This requires comparing identical sources, which in turn means you must capture the same performance onto two different computers at once using a splitter after the mic preamp, and with the same model sound card. Nobody can sing or play exactly the same twice, and moving even 1/2 inch between performances makes a very real difference in the frequency response captured by a microphone. When the differences are very slight it's also important to listen blind, 10 or more times in a row, with someone else switching so you don't know which source you're hearing each time.
Knowing how to properly compare audio devices and methods certainly makes one a smarter consumer, and learning even basic electronics helps you to separate reality from magical thinking and fantastic advertising claims. But it seems to me that accepting the frailty of our own hearing is the final frontier in understanding how audio really works.
[1] www.aes.org/events/119/press/pr_papers.cfm
Ethan Winer designs acoustic treatment products for RealTraps, and his new book The Audio Expert from Focal Press explains advanced audio principles and theory in plain English with minimal math. <www.ethanwiner.com>