(I'm from a country where earthquakes do not occur. For some reason, I always thought that earthquakes were accompanied by, well, movie-style earthquake sounds. It was kinda eye-opening to realize that earthquakes are silent...)
Where I live now, weak earthquakes can sometimes almost only be heard, along with a brief shake you'd easily mistake for someone slamming a door.
I was in a heavily affected part of LA during the Northridge quakes. It did 50b in damage, killed people, injured thousands. The most significant noise I remember was the cacophony of car alarms and people screaming.
Our house didn't collapse. If it had, i'd probably remember that noise.
Memory is faulty, so who knows. Maybe it was loud, but that's not what I remember, so I can understand why the descriptions vary so broadly.
No good footage of the Northridge event is really very available, but there is plenty of detailed (and unfortunately graphic) footage of the 2015 Nepal event, the 'sounds of chaos' far and wide out-match the sound of the low rumbling quake -- but that could be a deception of the microphones; I would imagine most of them lack the range to properly capture the low roar of an earthquake.
I think it probably depends on a lot of factors?
It has given me a lot of faith in the construction of my house though.
At the moment we have good results for predicting earthquakes (not forecasting) that should be able to warn the residents a few days before the impending major earthquake. The results are consistent based on offline data of the recent Turkeys, New Zealand, Indonesia and Philippines earthquakes. Hopefully we can reliably repeat the early detection capability with the real-time data, if fundings are available. If anyone know how to approach Japan government (or Turkish, NZ, etc) for funding the system please let me know in the comments.
[1] 1923 Great Kantō earthquake:
https://en.wikipedia.org/wiki/1923_Great_Kant%C5%8D_earthqua...