3 ms·

This effect is also used in dramatic movies, for example sci-fi and horror. It was used in the recent sound-award-winning Dune movie, where they used some inte

This effect is also used in dramatic movies, for example sci-fi and horror. It was used in the recent sound-award-winning Dune movie, where they used some interesting techniques for recording low-frequency sound, like recording moving sand dunes: https://www.youtube.com/watch?v=-KHcbp8szrY&t=257s https://www.youtube.com/watch?v=-KHcbp8szrY&t=257s There's also an app for detecting low-frequency sounds, although it seems surprising that cell phone microphones have the necessary frequency sensitivity. They seem to use some interpretation processing though the code is proprietary: http://www.toon-llc.com/support/lfd_en.html http://www.toon-llc.com/support/lfd_en.html

TeMPOraL 4y ago
> although it seems surprising that cell phone microphones have the necessary frequency sensitivity IANAAudio/Signals Engineer, but shouldn't microphones pick up everything down to 0Hz by default? For the stated sampling range, I assume the top end is limited by Nyquist theorem, and the bottom end by a low-stop filter in software.
- photochemsyn 4y ago
  Yeah, that sounds right. > "All current iOS devices include built-in microphones. These mics tend to be very consistent from one unit to the next, and are wide-range. However, Apple does include a very steep high-pass filter (which cuts low frequencies), presumably as a wind and pop filter. The low-frequency roll-off for the internal mic in these devices is very steep, on the order of 24dB / octave starting at 250Hz." However developers can turn the filter off.
- im3w1l 4y ago
  Normally you find the frequency content by using a discrete fourier transform. This takes a number of data points and transforms them from time domain to frequency domain. It's basically a mapping from one vector space to another. And the frequency domain vector space has fixed base vectors. The basis vectors correspond to frequencies, and being basis vectors they are actually independent of the data. What is data dependent is how much energy those frequencies have. A given window size will have a minimum frequency it can detect. So the low frequency cutoff is actually implicit. If you want to get lower frequencies you have to use longer windows, and that loses temporal precision, so you have to do a tradeoff. In a fascinating twist that is actually closely related to the heisenberg uncertainty principle.
- kortex 4y ago
  Not necessarily. A lot of analog front ends have a highpass filter around 8-20 Hz (the audible threshold). I don't really know why, it might be to prevent DC drift so that there is no asymmetrical distortion in the signal fed to the ADC, just speculating. But I do know a lot of systems just won't capture <20Hz.
- superkuh 4y ago
  The larger the wave size is compared to the physical aperture of the microphone the less efficient the surface is at converting the pressure wave into whatever (up to about 1/4 wavelength where things get weird). Electronics can make up for a lot of this to even things out but there are physical limits. It's reciprocal is more well known. To generate long wavelength sound efficiently you need large speaker apertures. In the same way you can't really use a 2cm speaker to generate bass you can't use a 2mm microphone to sample a 5 Hz (70 meter) pressure wave in the air.
Bayart 4y ago
On that subject, while I enjoyed Dune, I found the sound design so incredibly loud and annoying I almost left the theater.

TeMPOraL 4y ago

> although it seems surprising that cell phone microphones have the necessary frequency sensitivity IANAAudio/Signals Engineer, but shouldn't microphones pick up everything down to 0Hz by default? For the stated sampling range, I assume the top end is limited by Nyquist theorem, and the bottom end by a low-stop filter in software.

photochemsyn 4y ago
Yeah, that sounds right. > "All current iOS devices include built-in microphones. These mics tend to be very consistent from one unit to the next, and are wide-range. However, Apple does include a very steep high-pass filter (which cuts low frequencies), presumably as a wind and pop filter. The low-frequency roll-off for the internal mic in these devices is very steep, on the order of 24dB / octave starting at 250Hz." However developers can turn the filter off.
im3w1l 4y ago
Normally you find the frequency content by using a discrete fourier transform. This takes a number of data points and transforms them from time domain to frequency domain. It's basically a mapping from one vector space to another. And the frequency domain vector space has fixed base vectors. The basis vectors correspond to frequencies, and being basis vectors they are actually independent of the data. What is data dependent is how much energy those frequencies have. A given window size will have a minimum frequency it can detect. So the low frequency cutoff is actually implicit. If you want to get lower frequencies you have to use longer windows, and that loses temporal precision, so you have to do a tradeoff. In a fascinating twist that is actually closely related to the heisenberg uncertainty principle.
kortex 4y ago
Not necessarily. A lot of analog front ends have a highpass filter around 8-20 Hz (the audible threshold). I don't really know why, it might be to prevent DC drift so that there is no asymmetrical distortion in the signal fed to the ADC, just speculating. But I do know a lot of systems just won't capture <20Hz.
superkuh 4y ago
The larger the wave size is compared to the physical aperture of the microphone the less efficient the surface is at converting the pressure wave into whatever (up to about 1/4 wavelength where things get weird). Electronics can make up for a lot of this to even things out but there are physical limits. It's reciprocal is more well known. To generate long wavelength sound efficiently you need large speaker apertures. In the same way you can't really use a 2cm speaker to generate bass you can't use a 2mm microphone to sample a 5 Hz (70 meter) pressure wave in the air.

Bayart 4y ago

On that subject, while I enjoyed Dune, I found the sound design so incredibly loud and annoying I almost left the theater.