Twice the team, twice the Maya seats. $! Gustavo E Boehs Dpto. de Expressão Gráfica | Universidade Federal de Santa Catarina | http://www.gustavoeb.com.br/
2014-03-11 19:22 GMT-03:00 Raffaele Fragapane <[email protected]>: > There seems to be this mis-conception that benefits to small freelancers > are irrelevant to larger teams working on longer schedules and bigger > volumes. > > Of course the priorities of a place doing feature animation differ from > those of one producing MMOs, to those of a high end TVC boutique like the > Mill, to those of the individual hopping between 5 members rock-bands doing > 30 seconds skits. > > That said, there are good reasons, and considerable advantages, that are > shared across fields. > > If you look at something like brick-blur in the LEGO movie (objects > becoming a streak made of bricks representing large, real world volume > pixel equivalents past a certain velocity threshold) of course we could > have done it in another app. Parts of it towards the very end of it in fact > are in-house. But you know what? In the end it's practically a full > rendering engine that includes sampling options, bias adjustment and all, > and it was all done in ICE until the brick replacement and injection stage > that represents maybe 20% of the final effect. > Could I have done it in Maya? Yeah, I could, but for the same amount of > time I would have had a polished but really slow solution that would have > had mandatory flipbooks, instead of a 60fps brixel rendering engine running > in the viewport for animators to tweak in real time with controls > indistinguishable from the rig's own controls. > Could I have got it to run to 60fps in Maya? Again, probably yes, but I > would have had to manually and painfully write, tweak and debug some fairly > involved thread management, instead of being able to simply re-commit an > ICE graph that transparently updated for animators, and focus instead on > the creative challenges of nailing the effect. > > In the end ICE was preferred to both Houdini and custom solutions that we > had plenty knowledge and fire power to deal with had the need arisen. These > things add up, and they add up to the reason why Softimage has survived in > the rare film shop so long despite the added challenges of adopting a non > mainstream software. > > I've seen people genuinely surprised when they learnt that all the animals > in Life of Pi were handled by three riggers and one supervisor. Normally > that quality and amount of work would require more than double that crew if > you look at most credit rolls. > Well, Walking with dinosaurs was done with an average staff of 3.5 riggers > and one supervisor for its duration, and it had close to 20 unique species > and dozens and dozens of rigs once variations and ages are considered, with > 10 unique hero characters, and that's for a department that also took care > of a lot of conceptual work, creative iterations, simulations, and was > later migrated to take care of character FX. I think by the end of the > project the whole rigging department hadn't made it to the 100 hours of > overtime mark, and that's several people over two years. > > What do those have in common? Neither used Maya for rigging (Pi was > Voodoo, not Soft, just in case people don't know) :p > Had we used Maya several hundred hours worth of RnD and asset triage would > have been added to the bid, and the team would have probably have had to be > close to twice the size. >
