|
|
|
|
|
by frumiousirc
1567 days ago
|
|
I think this is simply using one rate for sampling and another for playback. Another comment suggests seismic samples like these are taken at 200 Hz. That is comfortably above the rock vibration frequencies of a few Hz. If one simply plays back these samples 100x faster the result would appear as 20 kHz sampling and the signal power that was originally around a few Hz would be heard at a few 100 Hz. The audio sounds a bit higher in frequency so perhaps these numbers are not precisely what were actually used but they give the idea. That's just the shifting. Accelerometers are used for the original sampling so there must also be some sort of pre-processing in the original sampling domain (the example 200 Hz time base) to convert acceleration information into displacement. This is a 3D displacement so then some other transform must be applied to represent the 3D vibrations with a 1D model. |
|