I have a BSP tree loaded into memory. First, I will find out on which sheet I am. Then I decode the PVS. Then I look at each area in this PVS and see if it lies in my truncation.
I was told that this is the wrong way to render, and that I really have to do reverse visualization by going through the BSP tree. How exactly am I doing this? I saw several different ways to do this, so I ask.
For example, rendering a Quake 3 BSP says: Go through the BSP to determine which sheet the camera is on. Extract and unzip the PVS for this sheet, iterate through the PVS and mark the leaves in the BSP. Pass the BSP near far. If the Node symbol is not marked, skip it. Test the Node Edge Box against the Frustrum camera. Add the current sheet to the change list.
Is this the standard way this is done today, or at least for simpler games.
source
share