Particles Local Orientation outwards - How to?

[Maximus]

Hello guys, i'm havin an hard time to find out how to achieve this effect..



I've setup this and now i want every single particle (instance) to point outwards locally, not all together.
I tried to modify the particle orientation but everything i do is done on all particles not on single also i cant find a way to have a proper orientation.

I am having this problem quite often and would be nice to create a compound that works once for all. Cause everytime i have to orient particles i have this issue.

For example, see how those points outwards based on their position?




Any suggestion?

Thank you

[Daniel Brassard]

Not in front of SI right now but, this would be a good application of the vector field node each vector pointing in the normal direction.

You will have to orient the triangles in the same direction as the vector field.

How do you define the triangle direction right now? (look to me they are all oriented in the -z axes)

Where is the pivot point relative to the two triangles?

[Maximus]

They dont have a custom orientation atm, that geometry has the pivot in the center. Its a normal grid and i melded 2 vertices in the middle to create an arrow-like shape.

[Maximus]

They all move together!
I have to try something different i guess.

[Daniel Brassard]

OK, the first thing i would do is determine the pivot orientation.

The Fibonacci algorithm give you a point in space. The point in space will determine where to put the triangle base location/pivot point (your are there from what I see).

The orientation is the angle between the origin point and the pivot/triangle base. The angle determine the direction of the tip.

So you could determine the vector direction from origin to pivot point and set the vector direction between the pivot point/triangle base to tip point with the same angle amount.

Basic vector mathematics.

Dan

[Daniel Brassard]

You have to do it for each instance/particle or they will move together as a group.

[Maximus]

well i think i got the math behind this, but its really hard for me to translate it into ICE nodes.
I would start with Get Data and dig into the "arrow" but i cant get what i would need. Sadly i think this is quite off my knowledge
Thank you for the help!

[Daniel Brassard]

If you don't get the answer by tonight, I'll put something together for you. You are almost there.

Dan

p.s. maybe this can help in between...

http://xsisupport.com/2011/04/11/ice-mo ... -to-edges/

[Maximus]

I think i solved it! gonna post in a few! Thanks alot!!

[Tekano]

splendid that you got it , was gonna say you are barking up the wrong tree with align particle by force or whatever compound that was - it may have been using the velocity vector. It was also a bit cruel of Daniel to leave you overnight stewing like that but to be kind, in order for you to learn, sometimes one needs to be cruel ;)

Ive been using ICE since beta and I still dont fully understand rotations or quaternions or the merits of splitting rotation up into quaternion or axis and angle or euler or rotation by orientation... its not exactly crystal clear on how to do this anyways with so many options and ways to connect and sometimes the factory align compounds don't behave as expected either!

but here is how I approached the problem. I knew I had to globally align to a position (the centre) then I had to rotate or tilt each particle around their axis locally. breaking it into 2 stages helped to try out various solutions and get each part working then it was just a matter of joining the two. heres my compound and screenshot, don't download unless you have solved it already!

this has got quite a bit of usage for those Motion Graphics folks too as you can alter the tilt by distance to null so the flower petals open and close - a bit like your reference picture where there is more tilt the further from centre

[Maximus]

Alright here is the final result- I was doing this compound for myself its based on some other compounds and some modifies i did so for anyone who find it useful here it is.



@Tekano, thanks for that elegant solution, i saw that it uses a Null, nicely done

[dwigfor]

This looks like a good exercise for me to try out. Glad you guys solved it.

Another useful node for rotations is Direction To Rotation. That's essentially a LookAt node (I saw that used a lot in XNA C# game code), which adds UpVector support.

[Daniel Brassard]

Me cruel ..... Never my intention!

Now for a bit of a challenge:

1. Scale the instance so that they are smaller closer to the center and get bigger the farther away from the center.

2. Rotate the instance so that they point to the center near the center and slowly point outward as the instance is getting away from the center.

3. Translate the instance so that they are close together at the center but slide downward (or upward if you prefer) on the main axis as they grow away from the center to simulate a plant growth.

4. Combine the Scale, rotate and translate into one effect.

(hint ... Think transform matrix scaling first, then rotate then translate in that order.)

5. Bonus point: add a bent deformation to the instance, controlled by the distance from the center, to add some interest to the shape.

Now I'm cruel......

EvilDan

[Tekano]

In anticipation of any of EvilDan's challanges, the compound posted earlier already takes into consideration his challenge points 1 & 2. For point 3 I had merely offset the centre of the instance petal so that its centre was on the bottom of the leaf. but it is worth including 3's functionality of an offset in Y based on distance for completion of Point 4.

Didnt get any bonus points for 5 though but for sure you should have a go and make it a feature rich 'particle localised orientation' compound

[Daniel Brassard]

I won't give up that easily, Tekano.

Challenge 6. Animate the growth of the particles, from the center outward following the path traced by the Fibonacci curves.

[Tekano]

growth... in what way the instance shape getting bigger?