View Revisions: Issue #5139

Summary 0005139: Renew particle declarations for rain
Revision 17.02.2020 17:01 by stgatilov
Description The current particle declarations for rain are flawed.
Aside from inherent problems, they are also not compatible with currently-implemented collisionStatic keyword.

Here is a brief list of proposed changes:
1)"deform particle" instead of "deform particle2" in materials
2) disable gravity, i.e. set "gravity 0"
3) drop distribution, i.e. set "distribution rect 0.000 0.000 0.000"
4) set constant speed, e.g. "speed 500.000"
5) remove "offset" --- I guess it is not necessary
6) remove "fadeIndex" and bump "color"
7*) remove "fadeOut" and set "fadeIn" to something like 20%
8*) set "orientation aimed" first number to 0.0

Without point 1 particle distribution is not uniform. This applies to ALL particles which are emitted from surfaces!
Without points 2, 3, 4 collisionStatic will not work, so should be fixed for rain.
Point 5 is optional.
Point 6 is questionable: it makes rain more visible even with less particles. Performance is not very good yet.

Point 7 fixes the stupid situation when all particles are always considered "faded", and allows to introduce proper fading in.
Point 8 improve performance two-fold. When the "aimed K" is set, then every particle is rendered as K+1 quads. Since rain goes in straight direction, curved trails are not a problem, so there is no reason to double quads number.
Revision 17.02.2020 14:41 by stgatilov
Description The current particle declarations for rain are flawed.
Aside from inherent problems, they are also not compatible with currently-implemented collisionStatic keyword.

Here is a brief list of proposed changes:
1)"deform particle" instead of "deform particle2" in materials
2) disable gravity, i.e. set "gravity 0"
3) drop distribution, i.e. set "distribution rect 0.000 0.000 0.000"
4) set constant speed, e.g. "speed 500.000"
5) remove "offset" --- I guess it is not necessary
6) remove "fadeIndex" and bump "color"
7*) remove "fadeOut" and set "fadeIn" to something like 20%.

Without point 1 particle distribution is not uniform. This applies to ALL particles which are emitted from surfaces!
Without points 2, 3, 4 collisionStatic will not work, so should be fixed for rain.
Point 5 is optional.
Point 6 is questionable: it makes rain more visible even with less particles. Performance is not very good yet.
Point 7 fixed the stupid situation when all particles are always considered "faded", and allows to introduce proper fading in.
Revision 01.02.2020 05:43 by stgatilov
Description The current particle declarations for rain a flawed.
Aside from inherent problems, they are also not compatible with currently-implemented collisionStatic keyword.

Here is a brief list of proposed changes:
1)"deform particle" instead of "deform particle2" in materials
2) disable gravity, i.e. set "gravity 0"
3) drop distribution, i.e. set "distribution rect 0.000 0.000 0.000"
4) set constant speed, e.g. "speed 500.000"
5) remove "offset" --- I guess it is not necessary
6) remove "fadeIndex" and bump "color"

Without point 1 particle distribution is not uniform. This applies to ALL particles which are emitted from surfaces!
Without points 2, 3, 4 collisionStatic will not work, so should be fixed for rain.
Point 5 is optional.
Point 6 is questionable: it makes rain more visible even with less particles. Performance is not very good yet.