Jump to content

fluid point instancing/Read from file - Mantra


Dries

Recommended Posts

Hi,

I'm doing an attempt to render a particle stream of water (who isn't doing fluids these days?) with metaballs and I need some optimization help.

I have generated gzipped .bgeo files for every frame up to say 10MB. They only contain point information.

The idea was to use a "Render from file" to reduce .ifd size, but I can't get "point instance" to work in this case. Is this normal behaviour?

I'm currently trying to copy the metaballs to the points and generate .bgeo's like that and then use "Render from file" but cooking time is extreme and my .bgeo's will go through the roof.

Before anything, I would like to visualize them just as spheres and I thought of the way you can render particle sysems "as spheres" but I tried to mimic the attributes that a particle system generates (pstate seems to be the only significat one?) but that didn't seem to work either.

Surely, somebody has tried this before...

Thanks

DRie.s

Link to comment
Share on other sites

Point instancing will not work. The metaballs won't interact as they are treated as discrete individual entities. You will have to send the metaballs as is to mantra.

Why not just save the geo file out as metaballs? You will have the weight attributes applied. Shouldn't be too heavy.

In H6.1, you can try out the procedural render dso for mantra to polygonize metaballs at render time.

Object>Render Folder

Geometry: Procedural

Mantra Procedure: + > choose Polygonized Metaballs

The native rendering of metaballs in Mantra is improved in H6.1. Try that first.

Before anything, I would like to visualize them just as spheres and I thought of the way you can render particle sysems "as spheres" but I tried to mimic the attributes that a particle system generates (pstate seems to be the only significat one?) but that didn't seem to work either.

Convert Meta SOP to polygonize your metaballs for previs purposes. It is fast and accurate.

The problem with using spheres is supporting all the different kernel functions and having a sphere that is the approximate size of the metaball. Sure you can generalize a solution, but it won't be too accurate, especially if you start randomizing the Weight and X, Y, and Z attributes with the different Kernel functions. Very tough.

Metaballs have pros and cons as far as the look vs. expense goes. They're great for morphing objects and doing effects like pouring water, dripping water, running water on surfaces, etc. They fall flat (pardon the pun) for water surfaces like a flowing creek.

A fluid simulator affecting a surface is quite common these days. Metaballs will fight you trying to simulate a flat flowing water surface. The metaballs have to be too small and you need too many of them to create this effect.

Now using metaballs to push a water surface around would make an interesting approach using bulge or magnet SOPs.

Link to comment
Share on other sites

Thanks for your reply.

I used metaconvert and saved the bgeo's like that, which helped, but as you mentioned, it looks as I need far more particles to make a substance. And even at low quality, the metaconvert is quite expensive.

I'm going to look at your "procedural" approach but I'm going to consider your surface transform option as well.

Cheers.

DRie.s

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...