Join Brian Bradley for an in-depth discussion in this video Why simulate and not animate?, part of MassFX and 3ds Max: Creating Simulations (2013).
When it comes to adding motion to objects in a 3D scene the question of whether we should animate or simulate may be one that we need to ask. Not of course that getting the answer will be simple or straightforward. The correct approach for any given project oftentimes comes down to the questions of time versus cost, versus benefit. One consideration, for instance, could be the number of objects that need to be animated in a particular scene or shot. If we all need, say, a single child's ball to come bouncing down the staircase and then come quickly to rest at the bottom, well a good animator should produce a fairly decent and believable result in a pretty short timeframe.
But if we need two hundred balls to come bouncing down the staircase, all with individual behaviors, all interacting and colliding with each other and the general environment, well, as you can imagine, that would take an animator a lot longer to create--if indeed it could be done with the required level of complexity at all. In such a case, simulation would probably provide us with a more suitable option. Another question that may well need answering could be, how realistic does the motion of my animated object or objects need to be? We may for instance be given a shot that involves a number of laundry items hung on a clothesline drying in a gentle breeze.
If this is designed as a cartoon-looking shot it is more than likely that our animators could pull this off quite convincingly using standard animation tools inside an acceptable timeframe. If, however, these items are to be seen in a photo-real setting, all of a sudden the believability of the cloth motion becomes supercritical. Given the complex and oftentimes subtle nature of cloth motion especially when affected by wind and other elements, the time required for an animator to attempt such a re-creation would perhaps once again make simulation a much more attractive option and will most likely give us a superior end result.
Time may also be a big factor when it comes to the question of whether to animate or simulate. More often than not these days, shots in production are given the shortest possible completion time frames. Because missing a deadline really isn't an option, it may will be that we simply don't have the time available to take the manual animation route, even though we may have the talent available to pull it off. And finally, unfortunate as this may seem to any artist who naturally wants to produce work to the highest possible standard, oftentimes it is the cost of a particular shot that will be the major consideration. Even if we do have the time available to animate a shot, the question of what it will cost in terms of man hours to dedicate one or more animate just to the production of a particular effect may well be the deciding factor in the end.
Time of course is money in the commercial world, and the ability to shave perhaps a number of days of the completion time for a project could make a big difference to its profitability. So whilst the question of whether to animate or simulate is, generally speaking, not going to be as simple or straightforward to answer as we would like, hopefully some of the thoughts in this video can just get us thinking in the right direction. If there are lots of objects in a shot that need realistic or natural motions and behaviors applied to them, or if we have a limited time frame and/or limited budget with which to work, well, we may want to seriously consider the value that can be added to a project if we use the simulation tools available to us.
- Setting up your 3ds Max project
- Understanding volume, mass, and density
- Learning the difference between concave and convex meshes
- Discovering Ground Collision and Gravity
- Baking out a simulation for rendering
- Adding an animation override
- Adding Rigid constraints and creating breakability
- Creating springy targets with the Hinge constraint
- Spinning targets with Twist
- Working with mCloth
- Putting a rip in mCloth
- Adding forces to a simulation
- Using fracture geometry in mParticles