Last time you learned about how pressure waves are converted into a neural signal in the cochlea in the ear. And you also saw that there was something missing from this process. Namely, that the sound wave after it entered the ear canal, lacked any information about where the sound came from. And yet we're able to localize where sounds are coming from. So how does the brain do this? That's what we're going to talk about in this video. So two sounds coming from two different locations, their sound waves would merge in the ear canal, and any information about where those sounds came from would be lost. But fortunately we're able to compute the location of sounds, and we do so in part by comparing the sound waves that arrive at each ear. And here's how this works. If the sound is located over here, it will arrive in this ear first, and be slightly delayed to arrive in that ear because of the longer path that it takes for the sound to get there. Even though the extra distance that the sound has to travel is fairly short, it's enough to make a measurable difference in how long it takes the sound to get there. Furthermore, the difference in path length varies smoothly with the exact angle. So a sound that is located more straight ahead has a fairly small difference in the distance of the far ear as compared to the nearer ear. A sound located farther to one side involves a bigger difference between the distance of the near ear and the far ear. These differences are referred to as interaural time delays, or ITD's for short. So let's do a few calculations. How long exactly are these delays? So to figure out those delays, we need to know the separation between the two ears. How far apart are the two ears? As well as the speed of sound. And we'll do the calculation for a sound located straight off to the right or straight off to the left, where the difference in path length would be the greatest, and so the difference in timing would also be the greatest. Well, I made a few measurements for myself. I measured the diameter of my head by taking off my glasses and measuring how far apart the two earpieces are. And I only had a ruler with inches on it, but I did some calculations. So, my ears are about 5 and a half inches apart. That corresponds to 0.14 meters. The speed of sound is about 340 meters per second. It depends a little bit on whether or not you're at sea level, as well as the weather conditions. That converts to 760 miles per hour. When I take that 0.14 meters difference in distance between my two ears, as well as the speed of sound, I come up with 0.00041 seconds. Time difference for a sound located off to the right, to arrive at the right ear before the left ear. So that's 0.4 milliseconds. That's an astonishingly short amount of time. Notice that that is less than the duration of a single action potential. Action potentials last about one millisecond, and can't occur more rapidly than about every two milliseconds or so. This is about a fifth of the minimum spacing between two action potentials. And furthermore, that's the largest difference that can occur. We're actually sensitive to much smaller differences. We're able to distinguish between a sound that is located straight ahead, and a sound that is located a little bit to the right, or a little bit to the left. So it's really pretty astonishing that our brains are able to detect these differences. And here's a little demonstration of what this sounds like. You might want to put on your earphones to hear this best. First I'm going to play a sample that has no inter-aural timing difference. No difference in the timing of the sounds that are delivered through the two speakers that form a stereo system. >> This should sound straight ahead. >> And it should sound pretty much straight ahead to you. For this one, the sound played through the left speaker will be about a half a millisecond ahead of the sound played through the right speaker. This should sound closer to your left ear. And this one is the opposite. This should sound closer to your right ear. It is thought that the way the brain does this involves delaying signals from one ear in comparison to the other ear, and having those signals converge on a particular neuron, that responds selectively to coincident inputs, as opposed to inputs that are coming in out of sync with each other. So here are two example neurons. So this neuron here, would be receiving input from the left ear and from the right ear, but the axon from the left ear and the axon from the right ear, might be of different lengths. Having the axon have different lengths would then delay how long it took an action potential coming from one side to reach that neuron, as compared to an action potential from the other side. So think for a moment whether or not this neuron would be more sensitive to a sound coming from the left, or a sound coming from the right. Well I'll give you the answer to that in just a moment. But consider this neuron down here, it would have a shorter path on this side, a longer path on this side. So it would also serve to be sensitive to sounds coming preferentially from one side versus the other, but it would be sensitive in the opposite pattern to this neuron here. If you think it through, this neuron is going to respond better when the sound arrives in this ear first. And the sound arrives in this ear second because the se, the action potential has farther to go on this side than on this side. This neuron would respond better when this ear heard the sound first. And this ear heard the sound second. While intraaural timing differences are not the only game in town, we also compare the sound loudness across the two ears to infer the direction that the sound is coming from. And here's how this works. So if there's a sound located off to this side, it's going to be louder in this ear. This ear is going to be in a kind of shadow. And the sound volume, when it reaches that ear will be a little bit quieter. And by comparing how loud the sound is in one ear compared to the other, the brain can infer the direction that the sound is coming from. So what does this sound like? This sample plays the sounds of the same loudness through the two speakers. >> This should sound straight ahead. >> For this one, the sound coming through the left speaker will be a little louder than through the right. >> This should sound closer to your left ear. And again, the pattern is reversed, here. >> This should sound closer to your right ear. >> But there's a critical problem with both of these cues. And you might want to pause the video here for just a moment, and see if you can guess what the problem is. Well the problem is called the cone of confusion. And that is that there're many locations that can produce the exact same timing and level differences. So if the sound is located all the way to the right, or all the way to the left, that will produce the maximum level difference signal. But for locations that are not quite all the way to the left or to the right, there're going to be many locations that have the same difference in path links and same head shadow effect. So any location on this circle or cone, will produce the same timing and level difference value. Any location on this cone- Will have the same timing and level difference value as any other location on that particular cone. This cone is different from this cone, but any location within this cone can't be distinguished on the basis of timing and level difference queues. So in the next lecture I'll tell you how we solve that cone of confusion, and do better than would be predicted, based on timing differences and level differences alone.