# How to generate evenly distributed random numbers from the tics of a Geiger Counter?

How can I generate random numbers from the tics of a Geiger Counter? Say I have a file with the timestamps for every event.

0
325
456
920
....


The time spans between two events can be described by a Poisson distribution.

What is the next step to get evenly distributed true random numbers from this?

One solution on Fourmilab compares the time span between two succeeding events and defines 0 and 1 by testing if the second time span is longer. This will create one bit from two normal distributed values.

Is there a standard method to create evenly distributed random numbers from the time spans more efficiently?