I just want to verify if what I will do is correct. Basically I want to save time by not “Building Light Cache” for each frame of my animation. If I can do this, I can probably half my render times.
Step 1:
Check Light Cache > Use Camera Path
Set Light Cache > Disk Caching > Mode > Single Frame
Set Light Cache > Disk Caching > Auto-Save / File to “mylc.vrlmap”
Step 2:
Render 1 image (Frame 0) to have the .vrlmap file generated
Step 3:
Set Light Cache > Disk Caching > Mode > From File
Set Light Cache > Disk Caching > Source File to the saved “mylc.vrlmap”
With V-Ray Next, the light cache is also used for other things besides GI (f.e. adaptive lights and the adaptive dome light) so while you might save some time on recalculating the light cache for each frame, you might lose performance in other places.
V-Ray uses the light cache pass to collect information about which lights are affecting which parts of the current view and adjusts the light sampling for the specific situation. This can save quite a bit of render time, especially if your camera flies through a complicated scene with many lights (f.e. rooms of a house). If you use just a single precalculated light cache, V-Ray will have to use some average sampling for all frames, which may be far from optimal and can actually lead to longer render times for each individual frame.