My friend told me this story from his antique radio club:
One club member is an audiphile and a former vibrations engineer for automotive companies. He disassembled his speakers and arranged custom housing for the drivers such that, based on his preferred listening spot, the peak of an average waveform from every driver would synchonize exactly at the spot where his ears should be. This, according to him, produces an unbeatable sound. We’re talking about opening a speaker and moving its tweeter, like, half a millimeter back.
No, I don’t understand how this is supposed to work, let alone consistently.
So I’m someone who actually designs and deploys speakers in large scale applications like arenas and performance halls. There is some truth to making all of the frequencies arrive at the same time based on your listening spot. However the “peak of an average waveform” is bullshit. The reason why time alignment is a thing is because of 2 reasons.
It takes time for the diaphragm that is moving the air that produces sound to move. A smaller diaphragm that produces higher frequencies will move sooner than a larger diaphragm that produces lower frequencies meaning the higher frequencies will arrive to your ear before the lower frequencies.
the speed of sound is not a constant. It varies with temperature, humidity, and atmospheric pressure. Unless you have a fully pressurized climate controlled space the speed of sound will vary, meaning the time alignment will vary.
Is any of this important? No. Sound (any wave really) will merge with other frequencies and produce a unified wave as long as the drivers are close enough together. So as long as the distance between the tweeter and the woofer is within half of the wavelength of the crossover frequency, it literally does not matter. The crossover’s between the tweeter’s and the woofer’s in most speakers are typically within 800-2.4khz, or 17”-5.6” (43cm-14cm) in wavelength. As long as your drivers are within that distance of each other the sound will converge.
And if it wasn’t, we can delay the electrical signal in the amplifier to make it all line up!!
As someone who designs audio equipment, I hate audiophiles, they make a mockery of my profession.
I don’t believe audiovoodoo but maybe he wants to achieve beamforming, which is a real thing, that’s how synthetic aperture radars work, as well as the latest Wi-Fi standards
I know you’re playing devil’s advocate, but to play devil:
In a theoretical world where you can manage to perfectly beamform the entire 20-20k Hz frequency range into a single node (or pair of nodes around the ears)… you’re still just re-condensing the original reference signal at the site of your beam target.
And if your idea of peak quality is to hear the reference signal loud and clear, it might be marginally easier to set up some well-tuned speakers in an arrangement relatively free of resonance hotspots and then crank up the volume.
So, how do you “crank up the volume” for that? Glad you asked; simple really: we need to apply a gain filter. To do this, we set up an array of batteries, and then connect only the positive side of the batteries to our audio cable. Positive electricity is bigger than negative electricity, so adding positive electricity to the cable means the speaker sound gets bigger.
In short, all you need to match the quality of a hi-fi beamforming speaker system can be replaced with a few 9V batteries connected to your tuner with a paperclip. Thanks for coming to my audio engineering TED talk.
That would work perfectly if he listened to music consisting of a single tone of different volumes…
Only way that could make any sense would be if he was trying to make sure the speakers were in phase. If you’ve ever had one wired backwards there’s an exact spot where you can get them to cancel out, it feels bizarre. But you don’t need to adjust the cones, just wire things the right way lol.
There’s no such thing as “the peak of an average waveform”, since it doesn’t make sense to average them.
They range from 20 Hz to 20000hz, so they have lengths between 2 cm and 1700 cm and peak everywhere in between.
What he was doing was attempting to avoid phase cancellation from stereo, which is an actual issue. You simply have to place yourself in the center, so that the distance to each speaker is the same.
It’s an issue for low frequencies, which have long wavelengths, carry a lot of energy and are usually centered (to avoid phase issues in the first place), so it is both possible and audible if they cancel out after leaving the speakers.
However, since they’re long, it also means that there is some wiggle room. Obviously there’s a perfect spot, but
It won’t have any noticable negative effect unless you’re like 1 meter away from that and accidentally sit in the perfectly bad position.
It is not as much of an issue for higher frequencies, since they have much shorter and more complex wavelengths, that are not necessarily centered in the first place. Even if they also theoretically do cancel out, it is unlikely that you would notice it happening at all, and if you do, you could simply move your head 1 cm to get into the right spot again.
In a car, the driver position is sort of fixed. It would make sense to angle the speakers for this purpose.
What he did doesn’t sound like it’s capable of achieving that. Unless he tried to make a bunch of separate chambers for different frequencies or something, you can’t do much with a single source (and we’re talking at least a few dozen resonance chambers, or else the sheer width of wavelength ranges makes this precision impossible). Beamforming is usually done with phased arrays of ultrasonic transducers, and measurement of how audio echoes in the room.
Might have been trying to make directional speakers, otherwise (transducer arrays are often used to make that!). But that’s ALSO not doable with such a small change, you need a bunch of bullshit like at least a cone, or once again by using more sources. Possibly he just modified the angle of the elements in the speaker to all aim at his head, which would actually be reasonable (yet not have much effect on quality, just volume), but it doesn’t sound like that’s what he did.
Could also simply put up some lazy ass half domes behind the speaker and behind where he would be sitting, but that’s ALSO not what he did
My friend told me this story from his antique radio club:
One club member is an audiphile and a former vibrations engineer for automotive companies. He disassembled his speakers and arranged custom housing for the drivers such that, based on his preferred listening spot, the peak of an average waveform from every driver would synchonize exactly at the spot where his ears should be. This, according to him, produces an unbeatable sound. We’re talking about opening a speaker and moving its tweeter, like, half a millimeter back.
No, I don’t understand how this is supposed to work, let alone consistently.
So I’m someone who actually designs and deploys speakers in large scale applications like arenas and performance halls. There is some truth to making all of the frequencies arrive at the same time based on your listening spot. However the “peak of an average waveform” is bullshit. The reason why time alignment is a thing is because of 2 reasons.
Is any of this important? No. Sound (any wave really) will merge with other frequencies and produce a unified wave as long as the drivers are close enough together. So as long as the distance between the tweeter and the woofer is within half of the wavelength of the crossover frequency, it literally does not matter. The crossover’s between the tweeter’s and the woofer’s in most speakers are typically within 800-2.4khz, or 17”-5.6” (43cm-14cm) in wavelength. As long as your drivers are within that distance of each other the sound will converge.
And if it wasn’t, we can delay the electrical signal in the amplifier to make it all line up!!
As someone who designs audio equipment, I hate audiophiles, they make a mockery of my profession.
I don’t believe audiovoodoo but maybe he wants to achieve beamforming, which is a real thing, that’s how synthetic aperture radars work, as well as the latest Wi-Fi standards
I know you’re playing devil’s advocate, but to play devil:
In a theoretical world where you can manage to perfectly beamform the entire 20-20k Hz frequency range into a single node (or pair of nodes around the ears)… you’re still just re-condensing the original reference signal at the site of your beam target.
And if your idea of peak quality is to hear the reference signal loud and clear, it might be marginally easier to set up some well-tuned speakers in an arrangement relatively free of resonance hotspots and then crank up the volume.
So, how do you “crank up the volume” for that? Glad you asked; simple really: we need to apply a gain filter. To do this, we set up an array of batteries, and then connect only the positive side of the batteries to our audio cable. Positive electricity is bigger than negative electricity, so adding positive electricity to the cable means the speaker sound gets bigger.
In short, all you need to match the quality of a hi-fi beamforming speaker system can be replaced with a few 9V batteries connected to your tuner with a paperclip. Thanks for coming to my audio engineering TED talk.
I have nothing to add, i feel stupid now
That would work perfectly if he listened to music consisting of a single tone of different volumes…
Only way that could make any sense would be if he was trying to make sure the speakers were in phase. If you’ve ever had one wired backwards there’s an exact spot where you can get them to cancel out, it feels bizarre. But you don’t need to adjust the cones, just wire things the right way lol.
At some point one must have considered wearing the speakers on their head…
There’s no such thing as “the peak of an average waveform”, since it doesn’t make sense to average them.
They range from 20 Hz to 20000hz, so they have lengths between 2 cm and 1700 cm and peak everywhere in between.
What he was doing was attempting to avoid phase cancellation from stereo, which is an actual issue. You simply have to place yourself in the center, so that the distance to each speaker is the same.
It’s an issue for low frequencies, which have long wavelengths, carry a lot of energy and are usually centered (to avoid phase issues in the first place), so it is both possible and audible if they cancel out after leaving the speakers. However, since they’re long, it also means that there is some wiggle room. Obviously there’s a perfect spot, but It won’t have any noticable negative effect unless you’re like 1 meter away from that and accidentally sit in the perfectly bad position. It is not as much of an issue for higher frequencies, since they have much shorter and more complex wavelengths, that are not necessarily centered in the first place. Even if they also theoretically do cancel out, it is unlikely that you would notice it happening at all, and if you do, you could simply move your head 1 cm to get into the right spot again.
In a car, the driver position is sort of fixed. It would make sense to angle the speakers for this purpose.
Audio beamforming is a thing.
What he did doesn’t sound like it’s capable of achieving that. Unless he tried to make a bunch of separate chambers for different frequencies or something, you can’t do much with a single source (and we’re talking at least a few dozen resonance chambers, or else the sheer width of wavelength ranges makes this precision impossible). Beamforming is usually done with phased arrays of ultrasonic transducers, and measurement of how audio echoes in the room.
Might have been trying to make directional speakers, otherwise (transducer arrays are often used to make that!). But that’s ALSO not doable with such a small change, you need a bunch of bullshit like at least a cone, or once again by using more sources. Possibly he just modified the angle of the elements in the speaker to all aim at his head, which would actually be reasonable (yet not have much effect on quality, just volume), but it doesn’t sound like that’s what he did.
Could also simply put up some lazy ass half domes behind the speaker and behind where he would be sitting, but that’s ALSO not what he did