I think Holly was referring to the use of Mip-Mapping in Carrara rather than the original problem.
The problem with demonstrating the Mip-mapping is that it’s more of a memory and rendering efficiency feature than a visual feature.
If it’s working properly you shouldn’t see any visual difference (or not much anyway, I can show some variation with my extreme textures that are designed to produce moire patterns).
Where the effect would be experienced is in better memory management when the renderer starts using the textures. The Mip-Mapping should reduce the texture size based on the object size in the camera view thus saving on resources. This feature is already used for game engines where the mip-maps are pre-stored in the texture format.
So, if you’re doing a large scene and are pushing the memory limits of your computer, that’s when you’ll notice the difference, otherwise not so much. It should also smooth out the ‘speckles’ that you can get in an animation as pixels on the texture are crossed by the rendered pixels.
Holly, am I right that you are finding that 8.5 handles memory better?