Archive for 3D Graphics
Busy, Busy, Busy…
Posted by: | CommentsIt’s been a while since I’ve posted, things have been really busy of late, and I’ve been doing a lot of traveling. There are a number of topics and compounds I want to share, but between work and home-life demands I just haven’t gotten to cleaning them up and writing them up. However, I just picked up a longer-term gig doing more feature film work and less commercials, so I may be able to settle into a routine that frees up enough personal time to get to some of this stuff soon.
Maybe I can get a bunch of stuff online as a Christmas gift to the community – I’ll be “heads down” in November but might have time after that…
Topics include:
- Fury 2 – this point renderer by Exocortex is starting to come into it’s own, with cast and self-shadowing, particle replication at rendertime, a standalone and maya version, etc.
- ICE terrain compounds – a long term project which I’ve detailed here, I’d like to add in some adaptive meshing (most of it was built prior to ICE modeling) and a better fluvial erosion compound, and release the compounds to the community.
- Electrical and lightning effects – I have a very nice “toolkit” of ICE tools for creating electrical effects which I want to clean up and release to the community.
- A slew of “utility” ICE compounds and one-offs for doing all kinds of fun stuff… from converting temperature to color according to a blackbody and color cycling compounds to branching strand systems and procedural snow deposition, etc. All compounds which come from production needs which I’ve been squirreling away with the intent of eventually sharing them.
ICE default nodes
Posted by: | CommentsI’ve been slamming away at production here at Method NYC, but when Vladimir Jankijevic posted this image to the XSI list I had to take a moment to put it here. This is an image of all the nodes which come with ICE by default. Considering that many regular ICE users add about 100+ of their own custom nodes, that makes for a pretty crazy number of options. ICE and Nuke both appeal to me for many of the same reasons… both are well thought out and excel at being both powerful and fast tools for demanding work. Now, back to this gig…
Images
Posted by: | CommentsThis isn’t anything special, just a slapped-together polygonized ICE/slipstreamVX pointcloud I made a while back.
Verlet Spring Water Simulation
Posted by: | CommentsIt’s been done by better TDs than myself, but a question arose on SI-community so I decided to pull out and dust off a simulation I slammed together during green lantern in a couple of hours, as an example scene. I put this together when we were doing rapid look dev for a “green energy” effect at PLF, where a woman is rescued by splashing into a surface of “water” which then turns into a flood and whisks her away from the fight…
You can find the compounds and the scene used to make this image here.
CG Society article on Green Lantern
Posted by: | CommentsCG society has an article about some of the work we did at PLF for Green Lantern. (I later went from pre/post vis to work on effects for the film at another studio, getting a double dose of green.)
More compounds for download
Posted by: | CommentsI admit it. I’m bad about remembering to post compounds for download. But not today… I’ve added a bunch of simple-but-useful compounds, some of the ones I use fairly often for a range of tasks. Included are a bunch of emission filters which can be used for all sorts of things, such as controlling placement of scattered objects, such as the tree cards in the images below. Enjoy. – AM
Fury Particle Renderer gets Furious-er
Posted by: | CommentsAs long as we are talking about Exocortex, they just posted this exciting preview of the next version of their point rendering tool Fury.
For those of you wondering why this is important, it’s simple enough. Fury is fast. Really, really fast. And it was written by Ben Houston, the original author of Krakatoa, a tool of choice for rendering particles. Softimage, Max and Maya users alike can move their simulations to ICE (or create their simulations with ice directly) and partake in the Fury awesomeness.
LOOK at it. 1 million points. Self shadowing and cast shadows. 1-second-per-frame.
“The major new features in Fury 2.0:
* GPU-accelerated particle self-shadowing
* Shadow maps
* Built-in compositing previewing.
* Command line renderer support.
* Synchronized Softimage and Maya support.
In this example, 1 million points are lit and rendered in about 1 second per frame and the shadow map is also created at the same time. Motion blur and DOF do not slow down rendering time.
The simulation in this example is from a alpha-version of SlipstreamVX 2.0 and thus the smoke motion isn’t quite perfect in this video.”
Congrats to Exocortex for Harry Potter VFX
Posted by: | Comments
There is a good article online about Exocortex and the work they did using Slipstream VX for the last Harry Potter film. You can find it here.
Some excerpts:
The pool, known as the Pensieve to Harry Potter fans, contains an oillike liquid that sloshes around in its basin whenever someone gazes into it. While they had managed to make do with available tools in previous Harry Potter movies, this time around animators wanted the effects to be bigger, better and more realistic than ever before.
Houston, 32, sold the company a piece of specially designed software that mimics real world physics and helps to accurately create water digitally. The water special effect had taken Houston more than three years to perfect. The software tool was designed to work with Gradient’s existing special effects software.
“I went and further customized our software to meet the needs of the Harry Potter production,” said Houston, adding the experience has been “awesome.”
Animators at Gradient learned about Houston’s software from a review in a computer graphics magazine, he said.
Having written the magazine article in question, I couldn’t be happier for Exocortex. Good work, guys!
GEAR – Use softimage? Get it.
Posted by: | CommentsJeremie Passerin is a TD and Rigger who has put a huge amount of time, thought and effort into an open source rigging toolset for Softimage, which he calls GEAR.
Very cool. But, you say, it’s for softimage and my stinking studio makes me use Maya. Well, Maya users, have hope. Jeremie has also been beta testing a Maya version, called MGEAR.
Let’s face it, there are more animators needing high quality rigs than there are high quality riggers. And rigging takes a lot of time. That’s why autoriggers came into being. If you are laboriously assembling all your rigs by hand and you just want to get to animating, you really, really need to be using a solution like Gear.
GEAR has the openness and enough well -thought tools to appeal to a dedicated rigging TD while also providing a fast and reliable way for animators to generate sophisticated rigs with a minimal learning curve. It’s modular in design, meaning it isn’t restricted to rigging only humanoids or quadrupeds, and it’s extensible, so riggers can use GEAR as a framework upon which they build in their own particular style.
It’s a very well thought out system and highly recommended for small studios and individual animators who need good rigs fast, as well as riggers interested in having the capabilities of an autorigging framework without the desire (or time) to make their own.
Mental Mill and Softimage
Posted by: | CommentsIf you are already an experienced shader writer, Mental Mill doesn’t have much to offer, auto-generated code won’t appeal. But for TDs who haven’t the experience or time needed to get a shader going, it’s a huge boon. It’s also a way to get a feel for shader code, letting you experiment and see how the code changes as a concept is changed.
Here’s the mental mill blog, with some info for softimage users.
What’s important for everyone to know about Mental Mill is that the same “tree” can be used to generate code for multiple rendering types – with a few caveats, you can create a shader for Mental Ray, a matching realtime shader, and a renderman compliant shader all in one go (ummm, Arnold? Oh well). For softimage users, this can give you custom solutions you can see in both the viewport and at render time. Here’s a video which covers creation of a realtime shader, for instance…
