http://www.sidefx.com/index.php?option=com_content&task=view&id=1694&Itemid=68

In terms of animation, Na’vi was most heavily relied on motion capture, and at early stage of production Weta Digital provided the basic mocap rigs (both facial rig for e-motion capture and body rig for rest of body motion capture) for principle characters to be enabled to be recorded onstage and be live-fed into James Cameron’s real-time virtual camera back in Lightstorm…but further fine tuning is always required by CG artists to clean up the data, and Weta Digital’s animators does step in to animate any gaps required to be filled in the areas where Na’vi interacts with environment objects as well as with other creatures. But in terms of overall performance and body language of characters are all kept almost perfectly intact by motion capture.

As for all the tiny emotive nuances of each Na’vi’s facial expression that we see in final result of film, we have to give a huge thumbs up for the facial rigging team and modeling team back in Weta for coming up with such a pain-staking pipeline to transcend each principle actor’s emotion so perfectly to the big screen:
http://features.cgsociety.org/story_custom.php?story_id=5501

When it comes to creature animation, motion capture data was used on horse (which turned into Direhorse in movie) that principle actors rode on, but most of other creatures were actually key-animated by Weta Digital’s animators (Viperwolves, Thanator, Banshees…etc.)

A lot of softwares and plugins/scripts are actually developed in-house that is only available within company as well, because so many tools that the artists needed to create Avatar more believable just didn’t exist in the current software market. Some of them may be released as commercial package in the near future perhaps (such as Massive crowd simulation software was developed from Weta Digital since Lord of the Rings, and Nuke from Digital Domain since Titanic…etc.), and some may never be shared and be kept as the company’s secret weapon to tackle extreme-realistic CG results

And just to further clarify what Larry said….Lightstorm came up with all the previz that was pretty much near identical to final film (in terms of overall character’s position and movement, rough jungle setups, camera movement…etc.), because back in Weta Digital or in any other company, we would always start off setting up shots accordingly on these previz that were given to us. James Cameron constantly refer back to these previz as we started chucking in character models and lighting scenes, since he’s gone through all these previz and ticked off approval back in Lightstorm already. However once in a while, certain shots will always change along the production in terms of composition as we further put in all the details, and those we have to experiment ourselves back in Weta Digital in shots such as Well of Souls sequence which has a lot of extreme color pallete to deal with that needed countless ongoing trial and refinement to get the right balance by the artists along with supervision of CG supervisors, Joe Letteri, and James Cameron to get it right. But the truth is because of these previz produced by Lightstorm was already approved by James Cameron….Weta Digital, ILM, Framestore and other companies involved, were able to compare and rely on them as a reference to get all these countless shots signed off on time.

All the companies involved in the Avatar have their own pipeline workflow (which usually takes years to be refined and continue to change from production to production) and each use their in-house setup to accomodate target goals and tasks given by the director, so however their technique and software use might be…so long they achieve their part of job, it’s mission accomplished, and a big thumbs up from Mr Cameron

So at the end of the day, what matters is all these softwares mentioned were needed to use in different stage of production to finish Avatar on time with the best quality that James Cameron wanted to achieve in the given production budget and time. And most important of all, the hardwork and talents of all artists involved from different companies is the most valued and only through sheer amount of passionate direction and perfection driven by countless CG supervisors, Joe Letteri and James Cameron’s vision made this movie even possible to be achieved to this insane realism level.

Of course also notepad was used to make the rough draft of the script

Lightwave was used to model real time props,good and then?,props can be made in max,blender,imagine,turbosilver,sculp4d.

What separates lightwave from max or silo as far poly tools?

I really dont get the point.

I understant lightwave fans but it just funny,main work was characters and even in the stupid and remote case that they where all made in lightwave,who cares?

Artist behind rigs,Renderman shaders,comp,animation,lighters are the real tools.

And again sorry but Weta uses Maya,for poly props like i said lightwave max silo have better tools.

But people where amazed by chracters not by the box in the background,and i could guess that for complicated props zbrush deserves more credit than lightwave

I wont bother again lightwave fans