GUIDED SCENE
Compositing
particles Binary
Loader 2: Modeling the stream
Vid1: Like a Rolling.... errr.. Water Stone?
This tutorial will show you some emission
tricks and how to set up particles with object interacion.
What we want to do, is create a ring of fluid that we will use to build
a sphere in the next tutorial..
1.- The first thing to do is to decide
wich emitter to use. So, let's see how we can achive the effect.
Let's take a look at the emitters list... Oh,
well there is no growing sphere emitter, so we will need to find out
how to do it:
We could use a sphere emitter, but is going to
look like a growing sphere, we are misssing the formation. It
would be the same as scaling a spehre... that's too simple.
We could animate an emitter moving around an imaginary
sphere and freeze the particles using the speed daemon (Max=Min=0)...
that would be cool, but not for this tutorial!
What about just a cricle emitter throwing particles
inside a spherical object so the particles make a loop defining
a perfect ring? Then we could use that ring several times from
different angles to complete the sphere.
Something like this:
This seems to be an easy task, but that's not going to be a perfect
sphere. As the particles are emitted from the circle emitters into
the sphere we are going to need something to remove those particles
from the mesh generation.... Don't worry, we'll get to that later.
2.-Setting up the particle loop.
As I don't want you to copy exactly what I do,
I'm just going to tell you how to do it step by step , and what it
should look like.
Lets get started:
Create a new project and add a circle emitter.
Bring in a spherical object. Since we are
not going to need the sphere for rendering and it's not going to
be modified, we don't really need to use a SD file. I'm going to
be using a 1 meter obj sphere.
The emitter is way to big or the ball is too
small. I'm going to increase the size of the ball. But how?...
scaling (from node), or using the global scale??...
I'm going to use the global scale (at top of the interface). I could
use the node scale, but I would have to type the scaling factor
for each axis.
The global scale will only affect objects so I don't have to worry
about the emitter, but that's not the most important thing. The
most important thing about using the global scale is that
RF is going to change the properties of the object to fit this
new shape, things like mass or collision distance will be set
to fit the new shape.
Okay then, I'll use the global scale to go for a sphere that's five
times as big.
You may also be wondering... Why didn't he scale
the emitter?. The answer is easy, becasue I like using the standard
RF resolution of 1. That will help to keep my numbers clear, and
just in case we are going to use fluid or gas, by using a resolution
of 1, we are going to get the standard fluid behaivor.. I'll get
back to this later.
Place the emitter so it's set like in the image above.
3.-Setting up interactions.
Go to the scene tree (F11) to make
the emitter interact with the object. Just press the 'add
all' button at the scene tree.
Press action and the particles will collide with the object.
bHowever, we need them to collide with the inside of the sphere.
We need to change the collision side
of the sphere
so the particles don't collide with the outside of the object
Select the sphere object and click over the particle interaction
tab. By deafult RF will set an object to collide using both
sides of a face. To change that, switch the collision side
to Inward.
4.-Saving out those particles.
Open the export central (F10). The export
central shows everything than can be saved with RF.
As there is only one emitter in the scene and
we want to save it, just press save 'All emitters'.
You will see that the emitter named 'circle01' is checked. Also,
a blue square will appear at the end of the time line, pointing
out that an emitter has been savde.
As particles are going to be saved, you will be able to play
the simulation back.
5.-Setting up the Emitter.
Press action. Now the particles
are going inside the sphere and colliding with the inside faces.
REMEMBER
Disable the viewport visibility
(Alt+D) to speed up the simulation.
Run the simulation and you should get something
like this:
If not, move the emitter to keep the emission inside the ball
from the beginning, but don't worry if some particles go outside
of the sphere weh the fluid completes a loop. Remember that
you can use the xform particles to move the particles
with the emitter
The emission is made by layers and if those
layers are too noticeable, we are going to get a bubbly mesh.
To avoid those layers we can add some variation to speed emission.
A good rule is to set the V random to half the value
of the Speed
6.-Setting up the simulation.
At this point there are two issues: The simulation
is not running fast, and the fx is taking too long.
Speeding up the simulation: This
simulation is really simple, there are no big forces in between
particles or daemons or object dynamics or other emitters. It's
just a fluid colliding with a ball, and we don't really need
the computation to be precise.
This is a good example of when the substep can be
decreased..
Since we are going to be playing with the subsetp, this is the
perfect momeent to save if you have not do so yet. Lowering
the substep will casue some instabiliy.
Go to the simulation options. There you will find the Max and
Min substeps which, by default, are set to 1 and 333 respectively.
Take down the 'Max substep' and hit action. Try using 100 instead
of 333. The simulation looks similar but now the calculation
is faster.
Try different settings but don't go extremly low or RF will
fry your computer! This simulation is so simple that can be
handled even with a max step of 2 (although it's not recommended
you do that). I'll we using a max substep of 20.
Speeding up the fx, This is looking right,
but the particles are traviling slowly. I want this to be 3
times faster. We could just increase the emitter speed, although
keep in mind that changing the speed will also change the fluid
behavior.
As we decided that we only want the simulation 3 times faster,
the fastest solution is to change the FPS (options panel) from
30 to 10. This means that the simulation is going to be the
same, but RF will not oputput 30 frames per second, just 10.
Run the simulation again. It should look like
this one:
6.-Removing the leak.
As the particles complete the loop, they avoid
the new particles to get into the sphere.
This means that the particles inside the sphere
are affecting the particles on the outside.
By activating the thin face test for the sphere(particle
interacion) the particles at one side of the object will not
affect the other side.
But this is not going to completly solve the leaking beacuse
there is too much pressure coming from the fluid on the inside.
We are going to have to delete those particles
using a destruction daemon; in this case a ~Volume Destruction
daemon. As there is just one emitter in the scene, any daemon
added will automatically be added to the emitter, so we don't
need the export central. Set the ~Volume daemon to Inverse
and place it where the leak is. (See the image below)
7.-The Final Touch. (optional)
The emission from the circle emitter is too thick.
I'm just going to use another destruction daemon, to remove
half of the emission.
Clone the actual volume daemon, set it
to interact with the emitter (scene tree), and place it as in
the following picture.
8.-...And the stream is done!
Run the simulation again. We are only going
to be using the first 120 frames, so you don't need to run
it until the end.