Especially this pic from och was an eye opener:
Both dashcams, both similar circumstances (no rain or snow whatever, just dark).
Sure the different headlights from the cars are the key difference... like lighting up that whole building in the back
The first sensor clearly has far less dynamic range in dark spots.
And I'm sure that the eye will see it differently again. Personally I've never seen a dashcam that can match the eye, and I had quite a few, also with the sony 323 sensor which is currently one of the better sensors in current dashcams when it comes to resolution/dynamic range at night (its a commercialized version of the IMX323LQN). I currently have one with the sony 323 and the 291.
This is a little bit conspiracy theory, but the police was very fast in stating that a normal driver couldnt have prevented the accident either (or at least thats what I understood from all of it). Thats a bold claim.
Are they (the police and possibly the political establishment) trying to protect Uber?
I mean on the one hand you have a dead victim. But homeless, convicted, a nuisance to society?
And on the other hand a large corporation. That has a very extensive testing program. To engage such a testing program, a lot of talking with government, politicians, police, experts has to be done to implement the specifics. Takes a lot of man hours. All have to be paid one way or another. Uber putting a lot of cars on the road. All with a supervisior. All paid jobs for people in the area....
Food for thought.