Realized I forgot the geometry term in transmissive bounce mult. 100,000 photons. 88 seconds

Fixed monte carlo. 10,000 photons, 1 bounce. 110 seconds

Final gathering. This sucked. 1 sample. I dont know how long it took but it felt like forever

Fixed specular mult component by adding geometry term. 88 seconds. Indirect + caustics. 100,000 photons

Same as before, but with quadratic flag set in EstimateIrradiance. 86 seconds

Increasing EstimateIrradiance search radius. 88 seconds. 100,000 photons

1 million photons. Indirect + caustics. 133 seconds

Caustics + indirect with photon mapping

Caustics with no other photon mapping so doing MC GI. 275 seconds

Direct + indirect with photon mapping. 1 million photons. 64 seconds

10 thousand photons instead of 10 million. Direct only. 41 seconds

Direct lighting only. 10 million photons. 310 seconds. 0 bounces