6) premultiplied alpha blending can merge batches with additive blending. traditional alpha blending cannot. note that dp's effectinfo stuff ONLY supports premultiplied blends.
yes, I should have added some dprints for the fallbacks.
13) for rain, "count 32" that means spawn a puff of 32 particles each time (allowing spark-like puffs), hence why there's a separate scaler.
for r_effect that spawns at the origin of the entity every single frame, then count is per second, while r_trail uses step instead, which is the distance travelled per particle.
15) clipbounce -1
0 will make it stop, 1 will slide along the surface (its a dotproduct thing), 1.5 will act bounce like grenades, 2 will maintain its speed, >1 && <2 will give you other momentum conservation values, -2 will make it become a decal (handy for blood), and -1 will remove the particle on impact.
if you also include a cliptype then you can cause the particle to spawn different types of particle on impact.
16) 'type static' is logically wrong. use 'type normal' (again, I should have added a dprint about fallbacks). with r_part, 'type' describes how the engine generates geometry, and is equivelent to 'orientation' in dp's particle system, thus 'type static' is as wrong as 'type smoke' was.
yes, I should have added some dprints for the fallbacks.
13) for rain, "count 32" that means spawn a puff of 32 particles each time (allowing spark-like puffs), hence why there's a separate scaler.
for r_effect that spawns at the origin of the entity every single frame, then count is per second, while r_trail uses step instead, which is the distance travelled per particle.
15) clipbounce -1
0 will make it stop, 1 will slide along the surface (its a dotproduct thing), 1.5 will act bounce like grenades, 2 will maintain its speed, >1 && <2 will give you other momentum conservation values, -2 will make it become a decal (handy for blood), and -1 will remove the particle on impact.
if you also include a cliptype then you can cause the particle to spawn different types of particle on impact.
16) 'type static' is logically wrong. use 'type normal' (again, I should have added a dprint about fallbacks). with r_part, 'type' describes how the engine generates geometry, and is equivelent to 'orientation' in dp's particle system, thus 'type static' is as wrong as 'type smoke' was.
Comment