Hearing Science Has Not Decoded Musical Enjoyment

Most arguments against higher definition audio claim scientific evidence showing that people can’t hear better than CD quality. They use this term “double-blind” that sounds powerful and authoritarian. AB and ABX tests are wielded like ancient tar bombs in debates, proving scientifically that no one can hear for shit! They refuse to believe their tests are lying to them.

They are blind testers because they don’t even acknowledge that we perceive music beyond our ears. They also don’t take into account spatial awareness, timing cues, and instrument timbre. They usually don’t conduct these tests in people’s known listening environment with their known songs.

A double-blind audio test is not a fair challenge, because the subject hasn’t tasted either before and neither tastes good to them. Given a short amount of time and unnatural switching, they pick their least hated.

Guess what, the people who control the study continually get what they want – confusion.

The DSP industry spreads FUD about true audio to further their business interests. That is to say people in the Digital Signal Processing industry (and there are lots of them online) spread Fear, Uncertainty, and Doubt to keep people listening and buying the latest digital audio trick.

If they can convince us we need and hear less they can give us less for more.

Nearly all tests that serve as the foundation for our hearing sciences are conducted with closed-ear headphones using artificial, singularly constructed test tones in some sort of lab or sound room. Rarely is the listener allowed to move about the room with speaker playback to find the sweet spot, in fact, many studies are done in mono!

First off – these tests are memory tests, not listening tests. Picking up the sound in real-time is very different than recalling the sound from memory. Since we cannot compare two complex sounds at the same time (like we can with images) we are left to compare a memory of a sound to a real sound. You can stop right here, this a fatal flaw in ABX testing.

There are fundamental differences between this and the way we actually hear and enjoy music, but these studies, in the name of science, try to remove all variables they can’t manage. This leaves us with a lab setup, and no one enjoys music in their lab. If no one is enjoying it then no one is hearing it in a natural way. The resulting data is garbage because the tests are garbage.

Sorry, people trying to decode this, good luck with your endeavors, but you are almost as blind as you are deaf to what’s actually going in in the world of vibration.

“The whole point of science is that most of it is uncertain. That’s why science is exciting–because we don’t know. Science is all about things we don’t understand. The public, of course, imagines science is just a set of facts. But it’s not. Science is a process of exploring, which is always partial. We explore, and we find out things that we understand. We find out things we thought we understood were wrong. That’s how it makes progress.” – Freeman Dyson, 90, Mathematical Physicist

Here’s a great article explaining several other reasons an AB test fails for music.

I believe there are several major flaws in the way AB tests are performed:

First Flaw: Memory recall of sound is very different than actual sound entering the body. Memory recall of sound comes from a different place in the brain than sound processing occurs. Memory storage is physically further away from the emotional centers of the brain than sound processing centers of the brain. Since we cannot compare two sounds in parallel (side by side) we are left with comparing actual sound to the memory of sound. This is not a valid comparison because these are two very different processes in our mind.

Second Flaw: 16 bit in real-time does not have enough space to carry all the data that our bodies can perceive. This has nothing to do with frequency range. This has to do with the depth of the data we bother to collect, digitize, transmit, then rebuild on the other end. To think the entire human ability to sense vibration can be recreated fully in 16 bit is like thinking that 8-bit video games were all humans would ever bother to build.

Google Earth has pretty good resolution these days. But can it see the blades of grass? Can it see pimples on my neighbors head? There’s always a higher digital resolution coming, because our actual analog resolution is infinite.

Third Flaw: The belief that suspending tiny speakers enclosed in a cup 1″ in front of our eardrums is equivalent to how we hear in the real world.

We hear through vibrations – not just vibrations in our inner ear, but vibrations picked up throughout our bodies. It’s proven that 2Hz to 6Hz is felt by most humans outside of the ears. It’s believed that when combined with the other frequencies of a complicated signal such as music we are able to deduce or feel high frequencies well over 20kHz.

A crash cymbal in person sounds different than a crash cymbal on mp3 or CD. A kick drum, a snare – hitting simple, ancient drums will expose digital audio’s flaws. Ever wonder why the high-hat went out of style? Because it sounds horrible in digital – the more you play it and vary it’s sound, the more the digital deficiencies show.

Perhaps it’s masking, or some form of sound science that we don’t yet understand, but anything they can’t measure from the inner ear is said to belong to “the sense of touch” as opposed to the “sense of hearing”. They just dismiss it and give it to another branch of science since they can’t quantify it.

If music plays and you are emotionally moved by it’s sound, are you hearing or feeling the music? I suppose both, mentally and physically. You mentally hear it, process it, and react emotionally to it.

Physically you feel it vibrating you, you feel it bounce through your ears, and you often feel a smile, tear, or other direct physical response from the power of the music, the power of the delivery of that particular song.

The other reason headphones are flawed is they can easily be radically adjusted. Simply moving them 1 millimeter any direction will produce drastically different sound, as will the state of your jaw and your mouth cavity.

Fourth Flaw: Test tones do not possess the complexity of anything close to music. Artificially produced tone doesn’t even get close to the variety and complexity of a single analog instrument. We have developed no scientific measurement for timbre or the multitude of other ways we hear and enjoy a sound. We seem to have very little scientific data regarding sound masking, sound blending, and sound character when it comes to our ears.

I’ve yet to come across a listening test that used as source material anything as complicated as a solo violin, an electric guitar, much less a symphony or your favorite p-funk album.

Fifth Flaw: Every room sounds completely different when playback occurs. Same source+ same song+same speakers+same amp+different room = totally different sound. Because of this we learn our primary listening environments and understand the sounds being created in those spaces better than in a strange room.

Believe it or not – you know, and have stored, what the rooms you usually listen in sound like. You know what your systems usually sound like in those spaces. You know the size and shape of the room, the makeup of the floor, walls, and other acoustic properties, even if you don’t think you care about this stuff.

You know if your system is malfunctioning even slightly, or if the overall sound of the space has changed for any reason. Our brains store ridiculous amounts of data about our environments beyond basic survival, which is partly why we are such an advanced monkey when it comes to building stuff to entertain ourselves.

Taking me to a strange place, playing strange things, and asking me to pick which is “better” will teach you nothing. Do it to 320 people and you have statistical mush. Having me relax in my usual environment, with my usual rig, usual ambient noise, usual mood and facial expression, usual everything except different source quality – that’s how you could do an accurate test of which sounds better. But that’s a human test, not a scientific test. So must continue to do this test ourselves and not wait for science to catch up.

Self Test: Do it yourself if you are a skeptic. Go to http://hdtracks.com and buy a song you already own in a lower format. Take both the files, duplicate them, and rename them random characters w/one different (this is to make you forget which one is which). Now sit there with your eyes closed and play the first file.

Listen close, listen entirely. Note that if you are playing from a stock computer/phone/ipod hardware, this difference might be small. Better plays show more obvious improvements.

Pay particular attention to depth, clarity, pan locations, cymbal trails, big reverbs, hard drum hits, kick drums, timbre (sharpness) on stringed instruments, particularly if bowed. Put yourself in the room where the music was recorded. Listen to that room in your ‘room’.

Either listen to the whole song, or stop after your favorite part and listen to the other file. Go back and forth during a good part of the song if you wish. Write down your initial thoughts, then keep listening. If you’ve never tried this before, your initial reaction could be the confusing – the ‘smaller’ file could sound louder and harsher, more direct, leading you think it’s bigger.

But if you get into the song, into the music and listen to all the parts, focusing on certain instruments or locations I’m almost certain you will end up picking the higher-resolution file for what it is. If you can’t hear a difference then keep practicing. Check hi-hats, cymbal crashes, echoes from lead voices for the clues.

If you just don’t hear any difference at all and don’t want to, then I guess consider yourself lucky? What’s out actually is “good enough” for you and you won’t need to spend any more money.

You might also want to consider listening to quieter music with a wide dynamic range (classical, opera, jazz, etc.) to rebuild your ears. Age and health have something to do with it, but ears can be trained and rejuvenated just like any other sense. You can learn to cook or enjoy good wine or feel something new at any age so don’t give up on your ears.

Bottom line: Putting me in headphones and playing beeps at different frequencies does not determine whether or not I can hear a difference in my music once data is removed from it.

16bit/44k “CD Quality” is 1978 digital technology, and some believe was compromised even in 1978 by marketing needs, and to stick behind incomplete science while defending any need to improve this after 37 years is a very strange phenomena. “Lossless” is just a marketing term to make 16/44 out to be better than mp3. But it’s still lost so much from the original masters.

Legendary record producers and legendary artists know masters sound better than CD’s. It’s just the digital babies defending their iPod’s convenience without understanding anything about what they are hearing, or missing. The dubstep affect? Loudness wars?

Or to explain the flaws of hearing science another way — check out one poor speaker reviewer’s take on how hard it is to measure sound quality.