StevenSurprenant
10-22-2011, 04:57 AM
In stereo reproduction, the sweet spot is equidistant from the two loudspeakers and set back to form an angle of 60 degrees. From this sweet spot, the listener can perceive spatial imagery at the speakers locations and also phantom images in between.
In multi-speaker reproduction, the sweet spot is roughly equidistant from all speakers. Phantom images are well defined between the front two speakers and the rear, but extremely fragile on the sides. What this means is that, (for this example we'll use a 4 speaker setup spread equidistant around the listener), if an image is panned between the speakers so that, conceptually it would go around the room, the phantom image would be easily located when it is between the front two speakers or the rear two speakers, but when this image is traversing between front and back or back to front, the image does not move smoothly between the speaker locations. In addition to this, the perceived image, to the side, has a preferential for the front speaker. As the image is panned from, let's say front to back, the image will “stick” to the front speaker during most of the time that the image is being panned and as the panning continues, we experience a loss of localization . What this means is that even though we perceive the sound coming from the front, our perception of where this image is in space will be less defined. At one point during this panning, a side image may form, but where it will form to the side is hard to determine. From there, a small incremental move in panning will cause the image to be perceived as coming from the back, but localization in space will still be indeterminate until the image is panned completely to the back speaker.
If the listener sits outside the sweet spot, the “precedence effect” will have an impact on the listeners spacial imagery. Normally we perceive the “precedence effect” as image shift in a 2 speaker (stereo) system. We've all heard it, but there is a little more to it. If the signal from one speaker is delayed between 0ms to 1ms, the image will shift closer to the nearest speaker (the speaker with the least delay). The closer the delay gets to 1ms, the closer the image gets to the closest speaker (or the speaker with no delay – same thing for our purposes). At 1ms, the image is perceived to be coming directly from the closest speaker. This assumes that the amplitude of both speakers remain identical to each other. This time delay can be created electronically or by simply moving one speaker further away. From 1ms to about 5ms delay, the image will appear to be coming from the closest speaker and the furthest speaker, for all practical purposes, will be undetectable. We can also say the same thing this way... The image will appear to be coming from the speaker whose signal arrives at our ears first. Once the delay exceeds 5ms, then what is called the “echo threshold” is reached and we begin to hear the further speaker again, but as an echo. The “echo threshold” can also be considered as the point where the “precedence effect” is being released. The fly in the ointment for these delay times I have given is that, sonic material with a more continuous content will change this threshold point.
To put this in a more realistic context, imagine that we have a typical 2 channel setup and that the speakers and the listener are all equidistant from each other and that the amplitude from each speaker is equal and remains so during the entire scenario- (Don't touch that dial!). Also, we send a signal to the stereo pair so that a phantom image is produced directly between each speaker. The listener remains sitting and listens while another person begins to move the right speaker further away. As the speaker moves further away, the listener will begin to hear the phantom image move closer to the left speaker. When the right speaker is about 1 foot further than the left, the image will appear to be coming directly from the left speaker. Now as the right speaker is moved further away, the listener hears no change, the image remains in the left speaker and for all practical purposes, the right speaker is undetectable. This continues until the right speaker is moved approximately 5 or 6 feet away at which point the “echo threshold” is reached and the listener begins to hear the right speaker as an echo. Keep in mind that this “echo threshold” varies with the content being fed into the speakers. The most interesting thing about the “precedence effect” is the range between 1 foot and 6 foot. The problem with using feet instead of delay time is that the larger the triangle we used when we set up our system, the further we have to move our speakers to achieve the delays required to hear this effect.
When speakers are further from a listener such as in a movie theater, the sweet spot is larger than compared to a typical home theater set up. What this implies is that the “precedence effect” is less prominent in a larger room, but what's important here is that room size (or speaker distance from listener) can greatly affect spatial imagery. A way to use this information is when you are setting up your system at home. The further away from the speakers you sit while maintaining an equilateral triangle, the larger sweet spot you have and the less image drift occurs when the listeners move laterally. You can also widen the sweet spot by simply moving the speakers further apart outside the equilateral triangle, but there are other issues created by doing this which is outside the scope of the subject matter that I am presenting at the moment.
At this point, I should also mention that this effect is the reason why surround systems uses a mono channel for dialog. A center speaker prevents image drift and so the dialog remains in the center of the screen regardless of where we sit. In other words, the “precedence effect” is eliminated from the center channel. The “precedence effect” still occurs in a surround system, but it is minimized in the front channels.
This is pretty much Audio 101 and there are many more aspects to consider if one's goal was to explain audio more in depth, but for my purpose, this is enough for now.
I hope some of you found this interesting.
Source information: Dr. Gary Kendall - School of Music and Sonic Arts, Queen's Univeristy Belfast
In multi-speaker reproduction, the sweet spot is roughly equidistant from all speakers. Phantom images are well defined between the front two speakers and the rear, but extremely fragile on the sides. What this means is that, (for this example we'll use a 4 speaker setup spread equidistant around the listener), if an image is panned between the speakers so that, conceptually it would go around the room, the phantom image would be easily located when it is between the front two speakers or the rear two speakers, but when this image is traversing between front and back or back to front, the image does not move smoothly between the speaker locations. In addition to this, the perceived image, to the side, has a preferential for the front speaker. As the image is panned from, let's say front to back, the image will “stick” to the front speaker during most of the time that the image is being panned and as the panning continues, we experience a loss of localization . What this means is that even though we perceive the sound coming from the front, our perception of where this image is in space will be less defined. At one point during this panning, a side image may form, but where it will form to the side is hard to determine. From there, a small incremental move in panning will cause the image to be perceived as coming from the back, but localization in space will still be indeterminate until the image is panned completely to the back speaker.
If the listener sits outside the sweet spot, the “precedence effect” will have an impact on the listeners spacial imagery. Normally we perceive the “precedence effect” as image shift in a 2 speaker (stereo) system. We've all heard it, but there is a little more to it. If the signal from one speaker is delayed between 0ms to 1ms, the image will shift closer to the nearest speaker (the speaker with the least delay). The closer the delay gets to 1ms, the closer the image gets to the closest speaker (or the speaker with no delay – same thing for our purposes). At 1ms, the image is perceived to be coming directly from the closest speaker. This assumes that the amplitude of both speakers remain identical to each other. This time delay can be created electronically or by simply moving one speaker further away. From 1ms to about 5ms delay, the image will appear to be coming from the closest speaker and the furthest speaker, for all practical purposes, will be undetectable. We can also say the same thing this way... The image will appear to be coming from the speaker whose signal arrives at our ears first. Once the delay exceeds 5ms, then what is called the “echo threshold” is reached and we begin to hear the further speaker again, but as an echo. The “echo threshold” can also be considered as the point where the “precedence effect” is being released. The fly in the ointment for these delay times I have given is that, sonic material with a more continuous content will change this threshold point.
To put this in a more realistic context, imagine that we have a typical 2 channel setup and that the speakers and the listener are all equidistant from each other and that the amplitude from each speaker is equal and remains so during the entire scenario- (Don't touch that dial!). Also, we send a signal to the stereo pair so that a phantom image is produced directly between each speaker. The listener remains sitting and listens while another person begins to move the right speaker further away. As the speaker moves further away, the listener will begin to hear the phantom image move closer to the left speaker. When the right speaker is about 1 foot further than the left, the image will appear to be coming directly from the left speaker. Now as the right speaker is moved further away, the listener hears no change, the image remains in the left speaker and for all practical purposes, the right speaker is undetectable. This continues until the right speaker is moved approximately 5 or 6 feet away at which point the “echo threshold” is reached and the listener begins to hear the right speaker as an echo. Keep in mind that this “echo threshold” varies with the content being fed into the speakers. The most interesting thing about the “precedence effect” is the range between 1 foot and 6 foot. The problem with using feet instead of delay time is that the larger the triangle we used when we set up our system, the further we have to move our speakers to achieve the delays required to hear this effect.
When speakers are further from a listener such as in a movie theater, the sweet spot is larger than compared to a typical home theater set up. What this implies is that the “precedence effect” is less prominent in a larger room, but what's important here is that room size (or speaker distance from listener) can greatly affect spatial imagery. A way to use this information is when you are setting up your system at home. The further away from the speakers you sit while maintaining an equilateral triangle, the larger sweet spot you have and the less image drift occurs when the listeners move laterally. You can also widen the sweet spot by simply moving the speakers further apart outside the equilateral triangle, but there are other issues created by doing this which is outside the scope of the subject matter that I am presenting at the moment.
At this point, I should also mention that this effect is the reason why surround systems uses a mono channel for dialog. A center speaker prevents image drift and so the dialog remains in the center of the screen regardless of where we sit. In other words, the “precedence effect” is eliminated from the center channel. The “precedence effect” still occurs in a surround system, but it is minimized in the front channels.
This is pretty much Audio 101 and there are many more aspects to consider if one's goal was to explain audio more in depth, but for my purpose, this is enough for now.
I hope some of you found this interesting.
Source information: Dr. Gary Kendall - School of Music and Sonic Arts, Queen's Univeristy Belfast