In this video, I'm going to talk about the rather surprising way that we can use sound frequency information to help us determine sound location. The clues to sound location that come from sound frequency are called spectral cues. So we've been talking about the cone of confusion that arises from using timing differences and level differences to tell us about sound location. Those cues provide information only about the horizontal position. We saw in the last video that movements can help resolve this, by moving your head, you can change the horizontal or vertical position of the head. Change the relationship of sounds to the head. And use multiple samples gathered from different positions to help figure out where the sound is located. In this video, I'm going to tell you about how we use sound frequency information to accomplish much the same end. First, I have to tell you what is sound frequency. So, recall, we made a graph earlier, of air pressure versus time. And if the sound wave looks something like this one, which is a sign wave, or a pure tone. Then it's fairly easy to describe what frequency is contained in this sound. And as you probably recall from physics or math, frequency is simply the reciprocal of wave length. But pure tones, that is sounds whose wave form constitutes a sine wave. Are actually fairly unusual in our environment. Most sound wave forms are more, are more complicated than that. And that's true even for sounds that are musical and have a definite pitch to them. So this is the wave form of a flute playing a particular note. And you can see that that wave form is not a clean sine wave. It definitely has a period to it. But it also has little jiggety jaggety stuff, in that wave form. So, how would you, would you describe the frequency of this sound? Well, that takes us to an analysis called Fourier Analysis. Fourier was a mathematician who realized that you could take any complicated wave form and deconstruct it. You could express it as the sum of simpler sine waves of different frequencies and with different amplitudes. So for example, this wave form here, has kind of an overall shape like that. Combined with a higher frequency, smaller oscillation to it. And I actually constructed that waveform by making this sine wave and this sine wave and adding them together. So Fourier analysis involves taking this complicated waveform and figuring out what simpler sine waves need to be added together to produce this more complicated wave. Viewed this way most sounds have multiple frequencies. These are three different musical instruments all playing the same note. Here's the flute /s, the pennywhistle, /s, and the banjo /s. So we can hear that those are the same note, but we can also hear that they are different instruments. We hear these as being the same note, because they have the same fundamental frequency. That is the low frequency large oscillations are the same in all three of these panels, but they differ in how many of these smaller high frequency components are present. Now we can graph this, and here's how we do it. So, let's take our example of a complicated wave form as the sum of two simpler sine waves. We can make a graph of amplitude versus the frequency of those sine waves. So here the low frequency component. Would be plotted here. It has a fairly large amplitude, so we'll give it a fairly large size on the y axis. This component is a higher frequency, so we'll plot it higher on the frequency axis, but we'll give it a smaller size because its amplitude is smaller. So, this kind of a graph is known as a spectrum. And spectrum provide a, a kind of a more complete picture of the multiple frequencies that can be present in a given sound. Okay, now a brief mention of how sound frequencies encoded in the auditory pathway. There are two ways that the auditory pathway accomplishes this. One involves the location of hair cells in the cochlea. And here is how this works. The Basilar membrane here has a resonance gradient along its length. It resonates or oscillates better to some frequencies than to others. And the precise pattern differs along its length. So, down at this end, it resonates more to high frequencies, and down at this end, it resonates more strongly to low frequencies. Hair cells situated at different positions will respond more to high frequency sounds versus low frequency sounds. The second way that the auditory pathway encodes sound frequency is in the timing of action potentials. Recall that as the Basilar membrane oscillates up and down, the hairs on hair cells are scraped back and forth, causing ion channels to open and close. The opening of ion channels is synchronized with the motion of the Basilar membrane. And the action potentials that they ultimately cause are synchronized with that motion as well. This synchrony is referred to as phase locking. All right but, what about frequency and sound location? Well, sounds are filtered by the external ear or the pinna, so everyone has little folds in their external ear and, those folds, alter the amount of energy at different frequencies. The filtering that they accomplish is direction-dependent. So different frequencies are attenuated to different degrees depending on what direction the sound is coming from. This frequency information about sound location is called a spectral cue. So here's how it works. Suppose there's a sound of a particular frequency coming from a particular speaker. That sound hits the outer ear, travels along some path. Bouncing, bouncing through those folds and then ultimately entering the ear canal. Maybe after bouncing in that particular pattern, it ends up inside the ear canal, slightly diminished in amplitude. A sound coming from the same location but having a different frequency might bounce in a slightly different pattern, resulting in a slightly different amount of attenuation when the sound enters the ear canal and so forth. So sounds coming from the same location and starting off with the same amplitude, end up in the ear canal having different amplitudes. [BLANK_AUDIO]. And similarly, sounds that have the same exact wave form, but are coming from a different origin, end up in the ear canal having been filtered differently and having different amplitudes, like this. So, the ear filters these two sounds. And then inside the ear canal, if we make a graph of pressure versus time, the signal from location A might be like this, whereas the signal from location B might be like this. These spectral cues vary in all directions. They vary horizontally, vertically, and with the front-back dimension. And they're used in concert with intraoral timing and level differences to help solve the cone of confusion. What I haven't yet told you about is what a complicated process it is to learn how to do this. It's something that we all learn how to do. And in the next video, I'm going to tell you how we accomplish this.