Why is that? I thought that it would be the same since it only depends on how long the camera shutters are on taking a sample, and how much things change?
ncastaneda02
Probability is just weird like that. The slide mentions it is a poisson distribution, which basically means that photons arrive at each pixel at some rate. Intuitively, if you were to wait the same amount of time for 1000 trials with the same underlying photon distribution, we would converge to an average sample, but from sample to sample we can get a decent amount of variance in the same way flipping the same coin multiple times under the same conditions will result in different results.
Why is that? I thought that it would be the same since it only depends on how long the camera shutters are on taking a sample, and how much things change?
Probability is just weird like that. The slide mentions it is a poisson distribution, which basically means that photons arrive at each pixel at some rate. Intuitively, if you were to wait the same amount of time for 1000 trials with the same underlying photon distribution, we would converge to an average sample, but from sample to sample we can get a decent amount of variance in the same way flipping the same coin multiple times under the same conditions will result in different results.