 |
Initial look shows the crate hitting the water twice, once with mass 21 and once with mass 9. Thus the two splashes, large and medium. So each instance is using the correct splash particle. The problem is why the code is detecting two collisions. |
|
|
When a moveable collides with water, the mass is adjusted based on how underwater the moveable is when the collision is registered. So the "mass = 21" event is when the crate is barely underwater, and the "mass = 9" event is when the crate is half underwater. |
|
|
The 2nd collision is registered 4 frames after the first. |
|
|
The movement code checks every 4 frames to see if the moveable has hit anything. If it finds that the moveable has hit water, then a determination of which splash particle to spawn is made. Since the movement code is finding the crate hitting the water in two consecutive checks, and the crate is still moving fast enough to warrant a splash, it plays a second splash. It has no knowledge of the first splash. This suggests a fix that sets a flag after the first splash occurs, and when the second splash is attempted, the set flag prevents it. The flag then gets cleared when the moveable comes to rest. |
 |
A couple thoughts: 1. Is this being affected by the uncapped fps change? 2. Sometimes multiple splashes are desirable--a floating plank, for example, spawns splashes at both ends. Also, some floating objects bob up and down and create small splashes while doing so, which looks reasonable. 3. If there are too many splashes it looks bad, because all the additive textures overlay each other until they're a pure white mess. This often happens with ragdolls, presumably because all the limbs cause different splashes. Maybe instead of clearing the flag after the object comes to rest, it could be cleared after 4+X frames, so the number of multiple splashes is reduced, but still allowed. |
|
|
Could you please commit this: textures/water_source/waterspray_m1add Edit: Nevermind, this just shows that I continue to be befuddled by texture and material files. Everything I need is there, but I'm not seeing any splash. Edit 2: I see the splash now. It's very fine, and if you're looking straight down on it, you can miss it. But it's visible from the side. |
|
|
1. No Looking at your suggestion. |
|
|
Source changes committed. Latest Windows binaries built and committed. Please test and provide feedback. |
|
|
Add a delay between the spawning of splash effects. Rev 7082: Entity.cpp Entity.h Liquid.cpp |
|
|
Resolved? |
|
|
My build has all kinds of graphics glitches since the last SVN update, but from what I could tell, the movable crate was not playing a splash particle at all. It plays a ripple particle only. Same with a ragdoll. Everything else I tested seemed correct. Vid in forum: http://forums.thedarkmod.com/topic/19042-looking-at-water-splashes/?p=411508 |
 |
Latest 32-bit SVN binary? |
|
|
If I throw the crate into the adjacent pool, I can see the splash (water_splash_large). Ditto with the body. What is the water setup at your end for that pool? |
|
|
I just tried with the default water_murk2 map on SVN. I had the same result. The water ripples portion of the splash particle played, but the splash stage did not. That confuses me a bit, since both stages are part of the same particle. water_splash_large seems to be the offender edit: I increased the particle count a bit and I can now see the splash effect, but it's extremely faint compared to water_splash. I'll play around with the particle and see if I can find the problem. |
|
|
Splashes look fine with your latest commits. |
|
|
Can this be closed? |
- Anonymous
