Mo2Chain Simple Fast and Parametric


#1

Mo2Chain Simple Fast and Parametric

Using only Cloners and parametric Objects,
no scripts, no fracturing, just out of the box.

QT Lookie 4mb

Scene file? I’ll put it up if the upcoming Chain tut at Cineversity is free

Cheers
Lennart


#2

Dang. You da man!


#3

Cool Lennart. You’re too polite, I wouldn’t worry about posting your file, you’re not stepping on any ones toes posting a scene file.


#4

looks good - except the weird way the links closest to the pin stay in a straight line… was that intentional?.. seems unnatural to me.

I like the shader too - like a clear coat… any insight on that? :slight_smile:


#5

Thanks guys.

@e[dub] Don’t you think it’s -more- weird that there is a pin floating
in thin air for no reason ? :slight_smile:
Na, joking, it was not intentional, that’s just how it looked when I ran the sim.
I guess stretching or slacking the initial stage of the chain would make a difference.
-Or-, maybe that’s how a chain looks? It gets stretch after all for that movement.

The color is just high narrow specular, reflection and transparency set pretty high
with a fresnel on them.

@muckymouse. The reason is the opposite, I’m afraid
(It’s not so much me being polite)
But I’d like to see the “official” solution (If i get the chance :slight_smile: )

Cheers
Lennart


#6

Looks very cool Lennart. I’m looking forward to see the scene file.

Peace,
Rich_Art. :thumbsup:


#7

Not quite sure what you mean. I’d say the more the merrier when it comes to ideas and files floating around.


#8

Great news Lennart :slight_smile:


#9

Using another shape :slight_smile:

Cheers
Lennart

QT Lookie


#10

Lennart, Isn’t the official solution to fracture?

Your solution looks vastly superior.


#11

I love this term “official” theres nothing official whatsoever about tricks. Maxon endorses their tools not how you use them, and as for Cineversity nothing there are the official ways to do it, heck we often teach several different ways to do things like the 3 different spiderwebs.

Fracture is only necessary in two specific situations, the first being when you have pieces or fragements you would like to split up that are all in the same polygon object. You need the fracture object for this because it has an explode segments option that will allow each segment of the model (groups of polygons that are not connected to each other) to be treated as if they were actually separate models.

The second is when you are procedurally breaking up a model like with the Shatter of polyFX or explosion and explosion FX deformers.

in the cases of separate objects you can use a fracture if you’d like or a cloner set to cluster. the advantage to a cloner set to cluster is that it allows you to have all those separate elements move as one object but get the more complex collisions, and several copies of that overall object.

See the following two examples (separate tutorial I’m doing than the chain one but same basic technique) In the first example I made the chair as a single polygon object and the collisions are funky because the objects essentially have a simple polygonal shrinkwrap around them and are one entire object. My options to solve this are to either model each piece separate (each leg, each support, the seat etc) and put them in the cloner as a cluster, which is fine if you plan ahead in your modeling, or, simply place them in a fracture objects with explode segments, which can treat them as separate objects and achieve the same effect as if I had modeled them separately.

Single Polygon Object

Modeled as separate elements

In many cases with more organic shapes, the separate objects won’t be a viable option, thats where the fracture comes in Handy, as well, fracture objects work with the connect object so you can weld your geometry together to make it behave like one cohesive object with proper phong, cloner cannot.


#12

Well official or not i’m curious - the chain problem is one of the first things I tried with mograph2, a solution that doesn’t involve exploding/fracturing complex objects would open up a lot of options. And if there’s a few ways of working it then all the better I say :slight_smile:


#13

Of course, we know that you guys aren’t producing the “official” solution. Its just a term for lack of a better word.

Lennart says he’s doing this w/o fracturing, even keeping parts procedural, and that piqued my interest. You gotta admit that sounds better than having to fracture and reconnect.


#14

In a simple object like this it is, the solutions offered by previous users are a global solution that can work in all situations, this method only works with simpler setups and with previous planning when designing your models. Without even looking at his first examples file he’s either sweeping or using sections of cylinders and torus (torii?)

In his second example it is clear he is using several sweeps Ask him to zoom closely on his curved areas though and you will see another limitation of this solution

The other thing is how it can complicate your setup often. Simple chains its is great for, chairs as well, shattering glass, previously existing models etc it isn’t, so its best to learn all. Keep in mid to it wasn’t us who posted the fractured chain.


#15

Without even looking at his first examples file he’s either sweeping or
using sections of cylinders and torus (torii?)

In his second example it is clear he is using several sweeps
Ask him to zoom closely on his curved areas though and you
will see another limitation of this solution

How close should I zoom? :slight_smile: What you see is a single sweep along the link.
QT ChainZoom

The chain and the concave shape problem in general can , and should most likely,
often be solved using the known methods for the reasons mentioned.
This method is not mentioned/used as far as I know. And since it is based on using
parametric objects only, I find it rewarding to be a little teasing, ok? :slight_smile:
It’s not fancy at all, it’s just looking at it from an other angle.

I think I do have a record of providing scenes/solutions for fun, good (and dumb) stuff
over the years. When time permits, I do like to get on the challenges
that pops up ever so often. I learn, we learn etc…
Maybe, this time, I just like to see some effort from others, based on what they see here,
to try ideas themselves before getting it “served”.

And Kai, put blunt but with love! Cheer up!
If replying in the first place, wouldn’t it be more positive to just reply with maybe a
“That’s interesting, wonder how?” or even “I’ll beat that!” instead of actually spend
time pointing out and exemplifying limitations of a solution.
Limitations there are of coarse, but not the ones you are stating as you can see.
Just a small thought. No offense, or hopefully I just totally miss your point!

Cheers
Lennart


#16

A challenge - argh! - and on a Sunday when I have work to do, how am I meant to meet my deadline now?!?


#17

Lennart, you are like a magician and if you take too long teasing the audience someone might figure out how you are doing it and blurt it out ;).

I thought of a solution a while ago that uses only parametric shapes too. There was a big drawback to my method though, slowness to calculate+viewport. But on the plus side it didn’t require any pre-planing.

Shall I wait for your official solution before revealing mine? :slight_smile: Maybe you found a better way – may I ask, how is your viewport performance?


#18

That I don’t mind at all (someone finding out that is) :slight_smile:

I thought of a solution a while ago that uses only parametric shapes too. There was a big drawback to my method though, slowness to calculate+viewport. But on the plus side it didn’t require any pre-planing.
Shall I wait for your official solution before revealing mine? :slight_smile: Maybe you found a better way – may I ask, how is your viewport performance?

Wasn’t it officially no “official” way ? :slight_smile:
The speed is no problem.

The magician comparison isn’t that far fetched thou.
There is a sentence in this thread that actually reveals the idea!

Cheers
Lennart


#19

Here is how mine worked.

I used a cloner to put primitives along the edges of my objects and use those as the collision mesh. So for each object, every edge has a cylinder running along it, essentially creating a wireframe similar to Atom Array.

:wink: This is how you are doing it, is it not? :lightbulb


#20

That’s clever and it’s not totally alien to how I do it, going to check
that out. I’m working with splines as base (the chain links), so it’s a bit different.
It’s family night now so I’ll have to get back to this later.

Cheers
Lennart