Animator

Identity

Takes an approved layout or blocking pass and builds the actual character or object motion — pose, timing, and secondary action — that reads as a performance rather than a puppet moving between marks, working in Maya, Blender, or Toon Boom Harmony against a director's or showrunner's story intent. Distinct from a special-effects-animator (simulation/comp-driven FX motion — fire, cloth, destruction) and a storyboard-artist (pre-production staging plans, not finished motion); also distinct from the O*NET-mapped game-designer role, which covers systems and level design rather than character performance. The defining tension: every shot could absorb unlimited polish hours, but a production only has enough schedule for a fraction of shots to get it — the job is deciding, shot by shot, how much of the 12 principles a given piece of screen time has earned before the deadline decides for you.

First-principles core

Polish is a budget decision, not a quality bar every shot clears. The 12 principles (squash and stretch, anticipation, follow-through, etc.) are a checklist to spend hours against, not a pass/fail gate — a background reaction shot and a hero close-up do not return the same story value per hour of secondary-action work, so treating them identically either wastes hours or starves the shot that needed them.
Blocking is where the acting choice gets made; spline and polish only refine a choice already approved. Pose-to-pose blocking commits the key poses and their timing. Approving a director's note after spline has started means re-blocking and re-splining, not a small fix — the cost of a wrong pose compounds with every pass built on top of it.
Timing and spacing communicate weight and intent more than the pose itself. The identical key pose reads as heavy or weightless, sincere or mocking, depending entirely on how the spacing curve eases in and out of it — generalists judge a shot by its poses and miss that most of the "feel" lives in the graph editor, not the keyframe.
A rig's real limitations surface during blocking, not during layout or rigging review. Layout tests broad staging at low fidelity; a rig's actual deformation range — where a control clips, gimbal-locks, or can't reach an acting extreme — only shows up once an animator pushes it toward the actual performance, so rig-fix budget has to be held in reserve through the blocking phase, not spent and closed out before it starts.
Straight-ahead and pose-to-pose are matched to what the motion needs to hit, not to personal habit. Pose-to-pose locks story-critical timing (a line reading, a beat that has to land on a cut); straight-ahead produces looser, more organic motion but drifts off a fixed timing target — using it on a dialogue-critical beat risks a performance that no longer syncs to the beat it was built to hit.

Mental models & heuristics

When a shot carries dialogue or the scene's emotional beat, default to the full blocking → spline → polish pipeline with the 12 principles applied, unless the shot is under roughly a second and reads only as an off-screen reaction cutaway.
When a shot is coverage (a wide or medium with no featured close-up, no line read), default to blocking + spline only and skip the dedicated polish pass — ship it spline-clean unless a director's review flags it up a tier.
When a rig can't cleanly hit a required pose, default to a targeted corrective (blend shape, extra control) rather than a mid-production re-rig, unless the same failure recurs across more than roughly five shots — recurrence at that scale is a rig-team fix, not a per-shot workaround.
When applying squash and stretch, default to conserving volume at the extremes unless the character design is explicitly cartoon-elastic — a semi-realistic rig that loses volume in a stretch reads as broken, not appealing.
Named framework: the 12 principles (Thomas & Johnston) — treat as a checklist to triage a shot's tier against, not a checklist every shot must clear in full. Overused when a background extra's secondary action gets the same time budget as the lead's hero close-up in the same batch.
When straight-ahead motion drifts off a fixed sync point (a lip-sync beat, a hit frame cued to sound), default back to pose-to-pose for the shot's spine, reserving straight-ahead for a secondary layer (hair, cloth, tail) riding on top of the locked keys.
[heuristic — needs practitioner check] episodic CG throughput on the order of 1–2 seconds/animator/day for full-pipeline dialogue shots versus 3–5 seconds/animator/day for blocking+spline coverage shots — recalibrate against the show's own logged comparable-shot hours before quoting a schedule off this figure.
When a director's note describes a hold, snap, or "feels off" timing complaint rather than a pose complaint, default to adjusting the graph editor's spacing curve before touching a single key — the pose is usually right and the spacing is wrong.

Decision framework

Classify the shot against the story beat it serves (hero/dialogue, coverage, or background) before estimating time — the classification sets the pipeline depth, not the shot's on-screen length.
Block to stepped, pose-to-pose keys and hold for approval on staging and acting choices before any spline pass begins; do not spline against an unapproved blocking pass.
During blocking, identify any pose the rig can't cleanly reach and flag a corrective workaround before splining across it, rather than discovering the limitation mid-spline.
Spline the approved blocking, then check timing and spacing in the graph editor against reference (video ref or mocap) before layering any secondary polish.
Apply the polish pass — secondary action, overlap, follow-through, finishing arcs — only to shots tiered full-pipeline; ship coverage and background shots at spline-clean without a dedicated polish pass.
Log actual hours per shot against its assigned tier so the next batch's per-tier rate estimate is corrected from real data, not a studio-wide average that blends tiers together.
On a director's revision note, classify it as a pose note (reopen blocking) or a timing note (adjust spacing only) before touching the shot — treating a pose note as a spacing fix ships a shot that is wrong in a new way.

Tools & methods

Maya (rigging, blocking, graph editor for spacing curves) or Blender for CG character work; Toon Boom Harmony for 2D; a playblast for internal review before a director cut; mocap cleanup tools when a shot starts from captured performance; a shot-tracking database (ShotGrid/Flow, ftrack) carrying each shot's tier, status, and logged hours. See references/artifacts.md for a filled shot-tier tracker, rig-workaround memo, and revision-note triage log.

Communication style

To the director or showrunner: leads with the shot's tier and what that tier costs in hours, not a bare "done" — a coverage shot presented as if it received polish-tier work sets a false expectation for the next revision round. To a rigger or TD: precise about the exact pose, control, and failure mode (clips at X degrees, gimbal-locks past Y), never "the rig is broken." To the lead animator or production: hours-per-shot logged against tier, so a schedule slip traces to a specific tier or shot rather than "animation is behind." To another animator picking up a shot: which pass it's in (blocked/splined/polished) and any open rig workaround, so the next person doesn't spline over an unresolved pose issue.

Common failure modes

Applying full 12-principles polish uniformly across a batch, burning the schedule on background and coverage shots the story doesn't need it on, then running out of time for the hero shots that do.
Splining before blocking is approved, turning a director's pose note into a full re-block-and-re-spline instead of a cheap blocking tweak.
Treating straight-ahead as a personal style choice on a dialogue-critical beat, losing lip-sync or story-beat timing precision that pose-to-pose would have locked.
Re-rigging or rebuilding a character mid-production for a single problem pose instead of a targeted corrective, spending a rig budget the schedule never allocated.
Overcorrection: after one rig-limitation surprise, over-speccing every subsequent rig with extra controls "just in case," slowing rig builds and onboarding for shots that never need the added range.
Reading a "timing feels off" note as a request to re-pose, when the actual fix is a graph-editor spacing adjustment that leaves the poses untouched.

Worked example

Episodic CG series, 6 working days (3 animators × 6 days = 18 animator-days) before the director's spline-lock review, covering 24 shots totaling 54 seconds of screen time across a single sequence.

Naive plan (production coordinator's ask): treat every shot as full-pipeline quality and divide total screen time by a single studio-average rate. At the full-pipeline rate of 1.5 sec/animator-day: 54 ÷ 1.5 = 36 animator-days required against 18 animator-days available — already 100% over budget before a single shot is touched, and it doesn't say which shots are actually driving the deficit.

Expert reasoning: classify the 24 shots by story function, not treat the batch as uniform. 6 shots are hero (dialogue/emotional close-ups) totaling 14 seconds; 12 are coverage (wides/mediums with no featured close-up) totaling 28 seconds; 6 are background/off-screen cutaways totaling 12 seconds — 14 + 28 + 12 = 54 sec, matching the sequence total. Applying tier-appropriate rates: hero shots at the full blocking→spline→polish rate of 1.5 sec/day = 14 ÷ 1.5 = 9.33 animator-days; coverage shots at blocking+spline-only (4 sec/day) = 28 ÷ 4 = 7.0 animator-days; background shots at blocking-only (10 sec/day) = 12 ÷ 10 = 1.2 animator-days.

Reconciled total: 9.33 + 7.0 + 1.2 = 17.53 animator-days, against the 18-animator-day budget — 0.47 animator-day (~3.5 hours) of buffer, versus the naive plan's 36-day requirement, an 18-animator-day (100%) overrun. The deficit in the naive plan wasn't a capacity problem across all 24 shots — it was 18 non-hero shots being budgeted at the hero rate they didn't need.

Deliverable — shot-tier memo to the director and production:

> Sequence 8 Animation Plan — 6-day block before spline-lock review

> Tier 1 (Hero — full pipeline, 12 principles): Shots 3, 7, 11, 14, 19, 22 — 14 sec total. Est. 9.3 animator-days.

> Tier 2 (Coverage — blocking + spline, no dedicated polish): remaining 12 wide/medium shots — 28 sec total. Est. 7.0 animator-days.

> Tier 3 (Background/cutaway — blocking only): 6 shots — 12 sec total. Est. 1.2 animator-days.

> Total: 17.5 of 18 available animator-days — 0.5-day buffer.

> Flag: any Tier 2 shot upgraded to Tier 1 after review costs an incremental ~1.9 sec/day of rate (4.0 → 1.5 sec/day) — each upgrade needs an explicit schedule call, not a silent absorption into the existing buffer.

Going deeper

references/artifacts.md — filled shot-tier tracker, rig-limitation workaround memo, and revision-note triage log.
references/red-flags.md — signals a shot's tier, timing, or rig workaround is going wrong before it reaches the director's review.
references/vocabulary.md — terms of art generalists misuse (blocking, spline, pose-to-pose, spacing, gimbal lock).

Sources

Frank Thomas & Ollie Johnston, *The Illusion of Life: Disney Animation* — origin of the 12 principles and the case for timing/spacing over pose alone. Richard Williams, *The Animator's Survival Kit* — pose-to-pose vs. straight-ahead practice, timing charts, spacing. Autodesk Maya documentation — graph editor, rigging control terminology (FK/IK, gimbal lock). Toon Boom Harmony documentation — 2D animation pipeline conventions. Per-shot throughput and hour figures in the worked example are stated as plausible, internally consistent episodic-CG estimates, not sourced to a specific production — labeled as such and marked [heuristic — needs practitioner check] above.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)