Idea: Ocean surface operator
#1
Hi!

Just an idea, but I think it would be very cool if tyFlow would have an operator which could apply ocean like patterns and movement behaviour to particles (f.e. using Tessendorf algorithms like this free plugin does: https://github.com/Perpixel/HOT4MAX/releases (even code is open source)).

With tyParticleskin you could then deform a plane with that particles (so it would be cool to use tyCaches with tyParticleSkin too).

The clue is with tyFlow you can easily test those particles against ship bodys or cliffs (colission, object test, surface test) ... etc. You coud spawn splashes or foam particles by testing particles of the ocean against some conditions (properties) or let break the waves. I imagine a bunch of super cool effects which could be achieved with such an operator.

I allready made a quick test (see below) using this plugin and recorded a tyCache from particles bound to that ocean surface.
In step 2 i applied the particles to a plane without that modifier via tyParticleSkin and the result looks great.
Unforunately tyParticleSkin doesn't work with tyCache so I had to re-input the recorded flow to a new tyFlow with BirthFlow and FlowUpdate, which is horribly slow, but I hope direct support of tyCache in tyParticleSkin is comming soon.
It would be so nice if tyFlow would have such ocean surface creating allgorithms natively implemented to create dynamic ocean surface particle patterns whose particles can be combined with all of the other mighty tyFlow tools and operators.

Thanks in advance!

   

   
  Reply
#2
I think ocean surfaces are better left to dedicated fluid simulators like PhoenixFD. Plus nowadays analytical ocean surfaces are often handled at a shader level, rather than a geo level.
  Reply
#3
Okay.

I agree, complex and physically realistic fluid simulations are best for absolute realism. But thats often not needed. The disadvantages are that they are hard to art-direct, because you get no direct feedback and the simulation often needs a huge forecasted frame range until you got the point where the interesting part of the simulation, which you need for visualisation, kicks in. This all takes several trys, much time, performance and space. So its not worth the effort in every case.

Shader based solutions are also good and deliver great results (f.e. poenix fd ocean tex and foam tex together with VRayDisplacementMod), but they are weak in interactions. Also heavy stormy scenrios with huge waves, breaking waves and splashing is not really possible with shader based solutions. Another disadvantage is, that you have to render a small sequence, again and again for tweaking and art-directing it.

The particle solution I imagine is somewhere between a full realistic ocean sim and a shader based solution. It should not be a replacement for a fluid sim and shouldn't have any fluid solver implementation. Just some lightweight pattern and movement creating allgorithms like the Tessendorf stuff.
The almost best advantage would be that it would be strong in interactions (in the range whats possible with tyFlow (and thats a lot Wink ) and can be used in combination with other operators for really cool creative stuff.

But hey, just an idea  Smile  !
  Reply
#4
Hi,

just to get a better impression I attached Maxfile and a short video.

Thats just absolutely basic, but I think you can imagine the possibilties that this would offer.

To open the maxfile you need the plugin from here: https://github.com/Perpixel/HOT4MAX/releases

To work with the second flow you need to create a disk cache from the first flow and feed it to the allready created but empty tyCache. Watch at the layers in layermanager, then it should be clear Smile.


.mp4   tyFTP_020-020-OceanPRTs_v1_0000_1.mp4 (Size: 6.71 MB / Downloads: 23)


.max   HOT_Ocean_Test_tyFlow_02.max (Size: 668 KB / Downloads: 7)
  Reply


Forum Jump: