surf´s up making of pdf

[thefreshestever]

after watching surf´s up again for the 2nd time, i wanted to know more about how they did the waves... found this, very interesting and quite impressive!

 

http://www.imageworks.com/ipax/docs/Siggra...CourseNotes.pdf

 

(Right Click and Save As)

[NancyGormezano]

That is a great reference!

[mouseman]

Agreed. For example, for the section on simulating different film types, they provide a list of a 18 attributes of aged film (e.g. Grain, discoloration, scratches, negative dirt (white), print dirt (black), etc.). Also, it is nice that they apply names for different things, e.g. "Ramped Speed shot. For these shots, the camera speed or the playback speed was adjusted during the shot for maximum impact."

 

Sure there could have been more detail or description in places, but it's very useful as is.

 

ETA: Another nice quote: "From the perspective of the Animation Director it was better to animate with a basic, stripped-down puppet that was fast, rather than use a slower rig with a lot of cool features. This way the animators could do many more iterations of a shot, and experiment with new acting ideas [...] The technology could not stand in the way of that creative process."

 

ETA: Just skimmed through the water sections. Wow, they did a lot of work on that. I'm in awe.

 

ETA: Page 24 - the wireframes look like splines/patches. Does Maya do splines/patches instead of polygons? It looks like they don't have 5-point patches.

[robcat2075]

ETA: Page 24 - the wireframes look like splines/patches. Does Maya do splines/patches instead of polygons? It looks like they don't have 5-point patches.

 

 

those are "sub-d" models.

 

Like A:M, sub-d stresses modeling the important contours of an object and having the computer fill the in between. In A:M that inbetween is the "patch", in Sub-d that in between is a gi-normous number of polygons. But at least you don't have to make each polygon yourself!

 

Unlike A:M, sub-d doesn't let you control the actual surface of the object, you have to manipulate a cage that influences what is inside it. NURBS have the same problem.

 

Sub-D is all-quads, only-quads , so it still has problem spots where we would be able to use a 5-point patch.

 

Here is a video that explains sub-d:

 

 

http://www.guerrillacg.org/home/3d-polygon...rfaces-overview

[brainmuffin]

Unlike A:M, sub-d doesn't let you control the actual surface of the object, you have to manipulate a cage that influences what is inside it.

 

That's some really out of date info there, Rob. The lines you see in that image (isoparms) are directly selectable and editable, as are the faces and vertices. The subdivision level tells the software how many polys make up each face.

 

The "problem spots" are called "poles" where five sides come together into one vertex. you can see one on Geek's cheek on page 24, it looks like a star.

[brainmuffin]

Here's a better example:

 

http://www.nevercenter.com/silo/videos/fea...subdivision.mov