Distance Attenuation Equation:
x = distance from the eye
Derived Size = clamp( pointSize * sqrt( attenuation(x) ) )
attenuation(x) = 1 / (a + b*x + c*x^2)
Default coefficients used are(they should work best with pointSize=1):
a = 0, b = 0, c = 0.000004f
This should give(without taking regards to the max,min pointSize clamping):
1. A size of 1 pixel at distance of 500 units.
Derived Size = 1/(0.000004*500^2) = 1
2. A size of 25 pixel at distance of 100 units.
3. A size of 2500 at a distance of 10 units.