The focus in games is to make software that is entertaining for our customers. Much of the problem solving in 3D is on the engine side.
On the content creation side, such as in Softimage, the focus is to integrate or mimic the engine or find ways of efficiently getting data to/from the engine. The critical part is to abstract game specific data so it is not tied to the content creation software, and inversely abstract the content creation software's isms from making it into the game engine. Do this while still providing an environment that artists can work quickly and efficiently. Not as easy as it sounds. The problem that is most encountered in Softimage and other 3D apps is you can't go low level enough to do what you need. Softimage doesn't really support custom classes and data structures as first class citizens in their API. Therefore, most implementations of toolsets are working around those limitations which often requires venturing into the dusty corners of the software where few people travel resulting in discovery of many bugs preventing you from reaching your goals. As for making games, it taxes your brain a lot more than film/video to figure out how to pull off an effect or implement an idea. Basically, your MacGyver skills are really put to the test. Matt -----Original Message----- From: Eric Thivierge [mailto:[email protected]] Sent: Tuesday, February 11, 2014 11:50 AM To: Matt Lind; [email protected] Subject: Re: Survey - how would you do this? Are all games made in an environment from what seems to be the early 90's? And true I probably wouldn't last long in games. I like using new technology too much. :) Eric T. On 2/11/2014 2:47 PM, Matt Lind wrote: > You wouldn't last long in games with that attitude. > > > Matt > > > > > -----Original Message----- > From: Eric Thivierge [mailto:[email protected]] > Sent: Tuesday, February 11, 2014 11:46 AM > To: [email protected] > Cc: Matt Lind > Subject: Re: Survey - how would you do this? > > With those restrictions, get a super fast animator to animate them by hand. > > Eric T. > > On Tuesday, February 11, 2014 2:23:31 PM, Matt Lind wrote: >> An artist came to my desk yesterday asking how to do what I felt was >> a simple task, but after getting 80% through it I ran into a speed >> bump realizing it needed custom scripting or other advanced tools to >> fully resolve to satisfaction. I had to give him a procedure that >> was ‘good enough’. This problem has multiple solutions, but I am >> curious how others would solve it: >> >> The problem: >> >> Artist must create an asteroid belt around a planet. The asteroids >> are likely 2D sprites which must face the camera and tumble as they >> orbit, but could be 3D objects as well. Asteroids must vary in size, >> shape, and animation speed (linear as well as rotational). Asteroids >> cannot collide with anything. Movement is generally slow – like a >> screen saver for your computer desktop. Asteroid positions are >> jittered within the belt. >> >> The question: >> >> Dispersing objects into a ring is fairly straightforward through a >> number of techniques, but how do you apply the random jitter to the >> object positions? >> >> The rules: >> >> -Cannot use ICE >> >> -Cannot use custom scripts, custom operators, or shaders. >> >> -Must only use tools out of the box that a junior or staff level >> artist would know how to use. >> >> -Must be able to create the asteroid belt, from scratch to >> completion, in less than 30 minutes – and be iteration friendly to >> react to art director feedback. >> >> -Ideally, the belt could be made a child of the planet in encompasses >> so it can be reoriented with respect to changes in the planet’s >> size/shape/tilt/orbit. >> >> -Final output must be able to exist with full integrity on its own in >> a vacuum. Cannot not have dependencies on custom code, external >> assets, or special case logic. >> >> -Asteroid belt fits within the default grid as seen in the scene >> camera. Think torus with diameter 40 SI units, and cross section of >> roughly 3 SI Units diameter >> >> Ready…..GO! >> >> Matt >>
