Also available as PDF
I've been meaning to write this blog for some time now, and I finally find myself with two days left in the year, and a house full of family (twenty-one people - "Oh My" indeed!). Sounds like the perfect excuse to hide away in my home office and write this blog entry.
I'll try to present some scenarios where splotches and blotches can occur and what steps can be taken to eliminate them from your renderings. To better understand what causes them, I'll begin with a short explanation on how Luxology's Global Illumination (GI) works.
There are two GI methods that can be used with MicroStation: Irradiance Caching or Monte Carlo. With Irradiance Caching (the default method), the render engine takes accurate samples for a portion of the pixels in the image then blends between them, producing the best combination of performance and quality. Monte Carlo methods sample every pixel at a reduced quality, frequently resulting in a grainier image.
To reiterate, the simplest way to think about Monte Carlo vs. Irradiance Caching is as follows:
Monte Carlo uses a lower quality (fewer rays) sample at every single pixel.
Irradiance Caching uses higher quality (more rays) samples at some pixels and blends them together.
Because of the way the Monte Carlo method operates, there will be significant variance from one pixel to the next - this is what can cause the grainy effect. Monte Carlo renderings must combat the grain effect by increasing the number of rays per sample, sacrificing render performance in the process.
On the other hand, Irradiance Caching will cause the variance to be spread between samples, possibly resulting in splotches when there aren't enough samples. Irradiance Caching offers a variety of options for improving quality while avoiding an increase in rendering time, either by increasing the number of rays used, firing additional rays selectively in problematic areas (also known as supersampling), or increasing the number of samples required to create blend (also known as interpolation values).
Now that you understand these basic terms and concepts, let's look at some real-life examples and the measures we can take to improve them.