Hi, I tried to study these two videos, but cannot follow
I can understand other related videos but these two.
Can someone help?
http://vimeo.com/50489166
https://vimeo.com/66237847
Any feedback would be greatly appreciated!
edit by owei: PELASE USE URL TAGS...!!!!
Triggered Deform on surface
Re: Triggered Deform on surface
I don't know that there is any simple way to explain it, but I'll try...
Look at the first video. There is a core concept here that he cleverly uses... I mentioned it in the update to the whirlpool deformer in the resource dump area here on si-community. You can use an inverse of an objects transformation matrix to calculate what the point positions of an object would be if it were at the origin.
In this case, instead of taking the transformation matrix of the object, Paul builds a transformation matrix for the object based on the point reference frame of the point closest to the null. When he multiplies all of the points of the sphere by the inverse of this matrix he is effectively moving those points such that the sphere is centered at the origin and rotated so that the "trigger" point is topmost. He then raycasts down to the face, looks at the length of the vectors which intersect the face, displaces those points by that amount, and then finally multiplies the new point positions by his constructed matrix to move the sphere back to where it was.
This general idea of moving points to be relative to the origin, doing something, and then moving them back is really useful, and not just for deformations.
For instance, I have a compound which moves particles in a pointCloud to the origin based on where a null is, imagine if the null was a parent of all the particles and you move it to the origin. I can then do interesting things, like measure the distance between the particle and the null on say the x axis and use that distance to color the particles.
In actuality I haven't moved the points, just calculated where they would be and used that to color them. The result is basically a directional light, any points close to the nulls x-axis get one color, say white, which falls off to another, black, the further those points are from the nulls x axis. You move the null around, rotate it etc and the particles color is updated correctly.
Hmmm. See what I mean? It's hard to explain simply, even though actually doing it is easy enough and very useful. If it helps, I've been meaning to try to tackle the subject in a blog post and I have some example scenes which may help when I do, but unfortunately I'm about to get busy on a big job and it may be 2 months before I can write anything up. Until then, look at the whirlpool example and see if you can puzzle it out, I believe I encapsulated the part where I move points to the origin and back in a compound. It will take a bit for the concept to become clear but when it does its really empowering. Cheers.
Look at the first video. There is a core concept here that he cleverly uses... I mentioned it in the update to the whirlpool deformer in the resource dump area here on si-community. You can use an inverse of an objects transformation matrix to calculate what the point positions of an object would be if it were at the origin.
In this case, instead of taking the transformation matrix of the object, Paul builds a transformation matrix for the object based on the point reference frame of the point closest to the null. When he multiplies all of the points of the sphere by the inverse of this matrix he is effectively moving those points such that the sphere is centered at the origin and rotated so that the "trigger" point is topmost. He then raycasts down to the face, looks at the length of the vectors which intersect the face, displaces those points by that amount, and then finally multiplies the new point positions by his constructed matrix to move the sphere back to where it was.
This general idea of moving points to be relative to the origin, doing something, and then moving them back is really useful, and not just for deformations.
For instance, I have a compound which moves particles in a pointCloud to the origin based on where a null is, imagine if the null was a parent of all the particles and you move it to the origin. I can then do interesting things, like measure the distance between the particle and the null on say the x axis and use that distance to color the particles.
In actuality I haven't moved the points, just calculated where they would be and used that to color them. The result is basically a directional light, any points close to the nulls x-axis get one color, say white, which falls off to another, black, the further those points are from the nulls x axis. You move the null around, rotate it etc and the particles color is updated correctly.
Hmmm. See what I mean? It's hard to explain simply, even though actually doing it is easy enough and very useful. If it helps, I've been meaning to try to tackle the subject in a blog post and I have some example scenes which may help when I do, but unfortunately I'm about to get busy on a big job and it may be 2 months before I can write anything up. Until then, look at the whirlpool example and see if you can puzzle it out, I believe I encapsulated the part where I move points to the origin and back in a compound. It will take a bit for the concept to become clear but when it does its really empowering. Cheers.
Re: Triggered Deform on surface
Thank you, I really appreciated for your explanation.
this is just what i need to know more.
And yes, I'm the one who ask about the updated whirlpool deformer in the resource dump area. but still hard for me to understand. my brain just don't function that way, so does for English...
I've try to read some articles, but yet to make the concept more clear.
I'll try my best, thank you.
this is just what i need to know more.
And yes, I'm the one who ask about the updated whirlpool deformer in the resource dump area. but still hard for me to understand. my brain just don't function that way, so does for English...
I've try to read some articles, but yet to make the concept more clear.
I'll try my best, thank you.
Re: Triggered Deform on surface
Don't give up! it doesn't come easy but you will get it I'm sure I struggled with matrices too I think everyone does when they get to that point
Re: Triggered Deform on surface
Hey, I've accomplished the first step, now i can make a wave go along with an animated null. looks like no different than making this in another way, because i can just get one... but it's a big change though. thank you.
But how can I make it goes with particles? and how can I trigger the animation base on the particles' age? How can I convert/switch per object to per point ? I remember see that somewhere, but I forgot...
any hint for these?
thank you
But how can I make it goes with particles? and how can I trigger the animation base on the particles' age? How can I convert/switch per object to per point ? I remember see that somewhere, but I forgot...
any hint for these?
thank you
Re: Triggered Deform on surface
ok, i kinda solved it, but still have one issue.
when new particle generate, the mesh-reading will change, then cause mesh breaking.see the pic.
dose anyone has any idea to fix this?
although I can just relax the mesh...
when new particle generate, the mesh-reading will change, then cause mesh breaking.see the pic.
dose anyone has any idea to fix this?
although I can just relax the mesh...
Re: Triggered Deform on surface
You have to layer them up, one after the other, not just use closest particle.
Re: Triggered Deform on surface
thank you.
i tried to make some changes and actually i don't know if my understanding is correct, i don't really know how to layer them up. just try...
now it becomes better but the breaking still happen in overlap, without overlap it's fine...
if i cannot fix it, maybe foam or turbulence will do.
i tried to make some changes and actually i don't know if my understanding is correct, i don't really know how to layer them up. just try...
now it becomes better but the breaking still happen in overlap, without overlap it's fine...
if i cannot fix it, maybe foam or turbulence will do.
Who is online
Users browsing this forum: brandwatch [Bot], trendiction [Bot] and 56 guests