It is often said that human hearing is incapable of hearing stereo sound at low frequencies. In this page I intend to argue that humans actually can hear stereo across the entire audible spectrum, but that in practice this degree of stereo separation is achievable only with headphones.
In Figure 1 we see a high-frequency sound wave passing around a human head. The sound is coming from the left speaker, and easily enters the head’s left ear. The size of the head (substantially larger than the wavelength of the sound) splits the wave, which doesn’t re-merge until some distance after it has passed the head. This causes the right ear to be in a “cone of silence” with respect to the sound coming from the left speaker. At this frequency, stereo separation is perfectly achieved; only the left ear hears the sound coming from the left speaker.
In Figure 2 we see that a mid-frequency sound wave flows more easily around a human head, and the sound from the left speaker does slightly impact the right ear. Stereo separation at this frequency is not perfect, but is still present to some degree.
In Figure 3 we see that a low-frequency sound wave flows effortlessly around the human head, and impacts both ears with full strength, regardless of what direction it came from. Stereo separation at this frequency is completely lost. This is why it is often said that human hearing cannot detect stereo separation at low frequencies.
Because of this effect, most audio systems include only one subwoofer, and its location in the room is irrelevant (from the standpoint of perceived sound direction) — no matter where it is placed, you will hear the subwoofer’s deep bass sounds as coming from all directions. This is sometime called the “omnidirectionality” of subwoofer-range bass.
When you wear headphones, the loss-of-stereo-separation effect described above completely disappears, and you get full stereo separation across the entire audio spectrum, because the headphones totally isolate the ears from each other. They deliver pure left channel to your left ear, and pure right channel to your right ear. This is why headphones have such a hyper stereo sound that can’t be matched by even the best speaker systems.
Is headphone stereo wrong?
The sound delivered by headphones can be called “unnatural” stereo, because you would never hear sound with this degree of stereo separation in any conditions other than wearing headphones. However, it is often the truest representation of what the artists wanted you to hear — if the artists want a sound to be audible by both ears (though perhaps to different degrees and/or at slightly different times), they can do that in the recording. But if a sound comes from only one channel, then presumably the artist wanted it that way.
Are unmodified live recordings devoid of unnatural stereo separation?
Not necessarily. If, for example, the left channel and the right channel come from microphones that were 200 feet apart, then low frequencies from an instrument or loudspeaker might easily impact one of those microphones more strongly than the other, just because of the different distances. To make a truly “natural” live recording, it might be necessary to use two microphones mounted on opposite sides of a roundish, human-head-sized, sound-dampening object. And then you would need to listen to the recording over headphones to ensure the same experience as being there — otherwise the loss-of-stereo effect would be additively repeated in your listening environment.
Some audio systems have a feature (I think Microsoft called it “spatializer”) that attempts to push stereo beyond what normally would be delivered by an optimal speaker setup (the green or red curve above). This is accomplished by playing a weaker, inverted copy of a soundwave out of the opposite speaker. For example, if the audio track has the left speaker playing a sine wave at 500hz, then the spatializer effect will play, out of the right speaker, a weaker 500hz sine wave that is 180° out-of-phase (i.e. inverted) with the one coming out of the left speaker. The idea is that some of the tone intended for your left ear gets to your right ear, but is exactly (hopefully) cancelled out by the spatializer effect coming from the right speaker.
It gets a little more complicated than that, since some of the spatializer effect from the right speaker is now reaching your left ear, so the left signal must be adjusted to compensate for that. This compensation reaches your right ear a bit, and so on. It’s an infinite regress, but a rapidly diminishing one, so it’s easy to calculate the needed strengths of each signal and get the intended results in each ear.
Even if perfectly executed, this effect would boost the stereo of only the mid-range frequencies (e.g. 500hz) since the high frequencies don’t need it, and the low frequencies don’t have enough stereo separation to permit the effect to work at all.
In practice, I think this effect would be very hard to execute correctly. A system that simply added an inverted copy of the entire left signal to the right signal (and vice-versa) would be a very bad implementation, as it would have uneven degrees of compensation across the mid-range spectrum, and would hurt the volume of low and high frequencies to which it never should have been applied. A really good implementation would have to run each channel through a specially designed frequency filter, before inverting it and adding it to the opposite channel, and even then it might require the effect generator to know the exact placement of the speakers relative to the listener’s head in order to create just the right frequency filter.
Would my music sound even more stereo if I used a spatializer effect while wearing headphones?
No; stereo separation would most likely be reduced. The ability of soundwaves to cancel each other, on which the spatializer effect depends, is lost when using headphones. Each cochlea gathers information about what frequencies are coming into it, then passes that information on to the brain. All information about wave phase is effectively discarded at the cochlea. Wave-phase-based effects, such as wave cancellations, must occur at the outer ear, and headphones completely isolate the left and right soundwaves, preventing the spatializer’s intended wave-cancellations from occurring.
Send me an e-mail!
Back to Tutorials.