So the plan is still to write up my notes on the best talks from GDC.  Until then, check out these free videos now available:

I thought it would be good to highlight some of the different workflow methods being used by top animators working today. Hopefully you can apply elements to your own workflow that will be improve your quality/speed & consistency. 


Name: Jason Ryan
Title: Supervising Animator
Studio: Dreamworks
Software: flipbook, Maya

Workflow:

• Planning : Uses 2d animation software to create golden poses. 
  
• Does not use film reference as he feels the animation will end up looking like him. 
  
• Flipbook is also used to define timing, flesh out acting choices and mood. 
  
• Final flipbook pass would be signed off by the director before moving into Maya.
  
• Blocking : Imports flipbook animation onto an image plane and recreates poses in 3d. Takes timing directly from flipbook. Makes necessary edits given context.
  
• Playblast in stepped.
• Sets keys to linear to see how the computer is interpolating between keys (without overshoots).
• Spline : No tweaking of curves in graph editor. Edits arcs and movement, tweaking poses, adding further breakdowns.
  

Name: Jeff Gabor
Title: Lead character animator
Studio: Blue sky
Software: Maya, video camera, video editor

Workflow:

• Planning : Thumbnails stage handled in Maya, creates poses to work out the relation of objects to the character (for interaction within a scene).  Also thinks about appeal and staging.
  
• Video ref : Films himself doing the scene, remembering the poses that worked when thumbnailing in Maya, camera & space considerations.
  
• Video edit : Chops up ref in video editing software, picks the best takes of an action, speeds up/stretches time to closer match timing of final shot.
  
• Blocking : Replicate final video edit by posing the actual model in 3d, timing taken from final video edit. Very conscious of how it will spline. Adds keys for holds.  Generally works on the rule of having a pose at least every 3-5 frames during motions.
  

• Splining : Works 15-30 frames at a time starting from the beginning. Works from the inside of the body out. Hips, Spine, Legs, Head, Arms, Toes/Fingers, and then face.
  Rarely tweaks tangents, just uses autotangents for moving holds and eases.
  

• Polish : Accepts he's lost his fresh eye by this stage. Gets other peoples eyes on it & asks what could be improved. Views a mirrored image of the anim to give that fresh eye again. 
  

Name: Keith Lango
Title: Technical Animator / animator guru
Studio: Valve
Software: GreasePencil, Sketch software, FrameMonkey, Maya

Workflow:

• Planning : Thumbnails, sketches out quick ideas on paper / pc
  
• Blocking : Creates key poses in Maya, working in layers. First layer – most important poses all keys on sequential frames. 2nd layer – keys shifted along in time line & filled in with most important breakdowns. A second layer of breakdowns is further added if he feels there still isn't enough information for the computer to calculate.
• Poses exported to Frame Monkey to work out spacing (although this seemed largely to do with sluggish frame rates in Maya on feature quality rigs around 10 years ago).
  
• After spacing is defined poses are imported back into Maya with same timing. Set to Linear.
  
• Spline : Only tweaks curves with viewport visible. Great advocate of not changing curves just to make them look pretty. If curve is funky but the movement is fine in Maya, he’ll leave the curve.
  

 
Name: Kenny Roy 
Title: Owner/ Director 
Studio: - Arconyx Animation Studios, LLC
Software: YouTube, sketchpad, Maya, video camera

Workflow: 

• Planning : Sets time aside to get used to the rig & what it can do. Sketches thumbnails. Records ref or grabs relevant videos from youtube. Acts out movement. 
  
• Blocking : Fundamental approach, stepped mode. Builds overlap into base key poses. Works with graph editor open making minor tweaks to familiarise himself with the keys in the scene. Even without dialogue, records an inner dialogue wav & uses it as a sound file in Maya, to help animate the feeling. Also adds 'performance texture' during blocking, so little shakes, bumps you know you definitely want in later. 
  
• Blocking plus: Turns stepped into copied keys to preserve holds. Adds further breakdowns to create arcs. Weight check pass. Compares poses to original thumbnails to make sure he's still on track. Reviews anim at full speed. Makes a list of notes from the first play of the anim, fresh eyes are the best critic.
• Polish : New mind set, polish is not an extension of blocking plus, all breakdowns should be there. Adds non- performance texture ( dirt like non- performance breaths etc) polish pass on arcs. Polish pass on weight, where is the center of gravity?. Then onto final 5%, toe splay, blinks, tiny tweaks.
  

Name: Tal Shwarzman
Title: Senior Character animator
Studio: Pixar
Software: Paper, video camera, video editing, photo editing, Maya

Workflow:

• Planning : Starts drawing thumbnails to help give ideas for the shot. They are not meant to be key drawings / frames or even extremes. They are just ideas.
  
• Records video ref  (Note: if you're not the right person for the reference, get the right person)
• Goes over the reference footage, picking out the most important poses. Then imports them & draws over the ref to push the poses.
  
• Blocking & Spline : Has settled on working on a mix of Layered and Stepped key workflow. Only works in what most software call spline or auto-tangent curve type, but still blocks in a pose based method.

I've spent a lot of time researching various ways to accomplish the same goal, creating solid believable animation. Although there are many different varying techniques out there they all really fall into two camps. Straight ahead, and pose-to-pose. Here's a very brief overview of what these two techniques entail.

Pose-to-pose

With this system, important poses for the complete action are concepted, perhaps even sketched via thumbnails before you even so much as look at a computer. From there those key poses are created (generally referred to as blocking out) in 3d, & from there timing is established. With the start & finish line clearly defined, important transition poses between these key poses are created, & then further transitions, until the computer has enough information to explain the movement believably.

Straight ahead

Although you may actually have an idea in your head, when animating straight ahead you do just that. You don't plan out your action or even create landmark poses for the complete motion. You may even move straight into spline & polish before even fleshing out the rest of the animation. The technique does created spontaneousness & you never really know what the end result can be.

So which one is better?

The answer (in my opinion) for the majority of us is a combination of both, dependent on the task AND the individual. Let me give you the pros of both.

Pose-to-pose

Using this system trains you to think about the complete motion. Although initially daunting (like staring at a blank piece of paper & hoping to create a masterpiece) once you have an idea blocked out there are less of those 'how on earth am I going to get from here to there' moments. It forces you to break down movement from reference, clearly recognise how body mechanics work & understand what parts of the body drive an action, what leads, what follows, what compresses, what stretches. It teaches you to create strong 'golden' poses, & visually makes your brain think of these poses more as individual drawings, instead of one great big muddy mess.

If you can master this technique, it also makes any type of action achievable. Using this system there isn't a reason why any type of movement cant be de-constructed, analyzed, blocked out & recreated. This analysis also helps build the confidence to create limitless types of movement, anything that is asked of you. By that I mean, when asked for a specific action, even when not knowing how it actually works, you KNOW you'll be able to de-construct it, because you use the same system to de-construct all movement.

Straight ahead

Sometimes you don't have an idea, you may have a mental block. With straight ahead that lack of an idea doesnt stop you animating. From experience I've started an anim without a clear idea of the execution ( but knowing broadly what the character needs to do, like fall over for example) & as I've started to animate, a few poses have peaked my interest & created sparks that have given me further ideas. With this approach you may try things you never would of attempted if planned out. This can lead to exciting pieces, but more often a lot of cul-de-sacs. The reality is, if you weren't prepared to try it out when planning, there was usually a good reason for it.

So surely it's not a combination of both? Pose-to-pose has way more pros than straight ahead !

True, but never forget technique is relative to the individual. This is where a combination really excels. You see every individual is different, they have different strengths & weaknesses. For example, if you have a great eye for seeing fluid motion, for being able to tweak on the fly & get things 'feeling' right, having weight, etc but your key posing isn't the best, then pose-to-pose will actually stifle you. It will actually make your animation worse! Why? Because your one main asset is moulding fluid movement in a layered fashion, & once you remove that & create static poses, you've lost it. You cant see it any more.

Just like the child who gets saddened by their awful drawing & no longer wants to draw ( & therefore never progresses & grows up hating art) you are left looking at your weakest ability (in this case poses) in all it's starkness. That will quickly depress you in much the same way as the childs drawing, & you'll give up on the anim before it's even begun. More importantly, you wont enjoy animating any more & this will also effect your results.

What I'm trying to say is, always look to progress, try to learn new techniques, but also play to YOUR strengths along the way. The reality is the pose-to-pose blocking approach will lead to a higher quality of animation if you can master it, but there's nothing to say you cant combine your strengths as well so that you improve organically, whilst enjoying the process.

I have worked with animators who really excel at the pose-to-pose / blocking out technique. They can break an action down & their blocking is incredibly complete. By the time they've finished this stage, they could just hit 'auto tangents' & the anim would STILL look pretty sweet. I've also worked with animators were their blocking leaves lots of questions unanswered & they have to add a lot more in the spline stage, because only at that stage can they see what's missing & how to plus the anim.