[Sidefx-houdini-list] particle fluid emitter & pop solver

jlait at sidefx.com jlait at sidefx.com
Wed Oct 24 14:16:31 EDT 2007


On Wed, Oct 24, 2007 at 05:22:03PM +0200, Abdul-Kareem Abo-Namous wrote:
> *both* POPnets implies that I should have two of them... I'm using one for
> the pop solver (and called the source null pop source1 as per default) to
> apply the POP forces (in theory), where should the other one be? As far as I
> can see, the Pop Flow example also has only one POPnet.
> 
> What do I need the other one for?

    The other one is buried inside the Particle Fluid Emitter.
Particle Fluid Emitter and Sink both have POP networks inside them to
do the actual particle generation and deletion.

    As it happens, the rough consensus we achieved was to call the
root of the particle system "location1" everywhere.  Unfortunately, it
seems I didn't update the Pop Flow example file so it still uses
"source1" as the name.
    The fix is thus to rename the source1 node inside the popnet to
location1.

    The esoteric reason for this is that whenever a POP net cooks it
needs to know where to start the chain.  Say you have a Merge Pop -
each particle needs to know which branch of the network to move along.
This is done by the generator primitive attribute.  The particles
created by the emitter get a generator attribute of "location1" and
then, not finding a suitable root node in the attractor pop net, do
nothing rather than cook properly.

    I'm also going to update the popflow demo to use a VOP SOP to do
the acceleration -> force conversion.
-- 
        - Jeff Lait




More information about the Sidefx-houdini-list mailing list