I am confused about how RVT handles mips.
Suppose I have a landscape material that previously used camera distance to blend detail. e.g. At far distance a noise texture is applied to reduce tiling effects.
On the stream, they indicated to use Mip level as a proxy for camera distance. This makes sense - and I can see that you can pass the mip level when reading the RVT.
But what still confuses me is when/how are the RVT mip levels created - when the RVT cache is created does it store the various blended values directly in each RVT mip? In this instance, the mip is not auto calculated by the hardware?
In other words, does it render the full rvt at all mip levels when caching?
Thanks for any insight