The Advanced Motion Designer's Guide to Generating Storyboard Frames with Consistent Character Poses Using Stable Diffusion

You are a specialist motion design director with 13 years of experience producing animated brand content and explainer video storyboards for agencies and in-house creative teams. Help me build a character description prompt so I can reduce pose inconsistency across multi-frame storyboard sequences. My situation: - Character role and visual archetype: [e.g., "confident mid-30s female founder, business casual, South Asian, medium build"] - Number of storyboard frames in the sequence: [e.g., "12 frames — 4 scenes, 3 angles per scene"] - Action or emotion range the character must convey: [e.g., "presenting, listening, reacting with surprise, walking away"] - Art style and rendering mode: [e.g., "clean vector illustration, flat color, bold outline" / "semi-realistic digital painting"] - Brand color palette the character design must work within: [e.g., "cobalt blue, warm sand, charcoal — no red or green"] - Output format and frame dimensions: [e.g., "16:9 at 1920x1080px per frame, exported as PNG"] - Biggest current failure in character generation: [e.g., "face changes between frames, clothing color shifts, hands render incorrectly"] Deliver: 1. A locked character description block of 80–100 words that defines physical appearance, clothing, and visual style with enough specificity to survive across 12 separate generations without drift 2. A ControlNet pose recommendation for each of the four action states — presenting, listening, reacting, walking — including which ControlNet model to load and what reference image type to source 3. A negative prompt string of 20 terms targeting the specific failure modes of multi-frame character consistency: face drift, clothing recolor, hand distortion, and scale inconsistency 4. A seed management protocol that specifies when to lock, when to vary, and when to fully reset the seed across a 12-frame sequence 5. A prompt structure template showing the exact order of character block, pose instruction, background instruction, and style tag for each frame 6. A hand generation sub-prompt — a dedicated 30-word clause that improves hand rendering accuracy when appended to any character prompt 7. A frame-to-frame consistency scoring rubric with five visual checkpoints to assess before the storyboard is presented to the director or client 8. A LoRA selection guide — three criteria for deciding whether to train a custom character LoRA versus continuing to prompt-engineer for consistency **Write every output as a technical production brief, not a creative brief — this storyboard will be used as a direct reference for an animator, so precision beats inspiration.**