Character Consistency Bible

PARITY — Production Board 03

How to make Kira Vasquez look like the same person across 14 shots, 7+ AI models, and 90 seconds of film.

The Core Challenge

Every AI model generates faces independently. Veo 3.1 will render Kira differently than Sora 2, which will render her differently than Kling 3.0. Without a deliberate consistency strategy, she'll look like 14 different women across 14 shots — and the film falls apart. Character consistency is THE hardest problem in AI filmmaking, and the entire industry acknowledges it. Our strategy uses Higgsfield's native tools (Soul ID, Popcorn, Kling O1 Element Library) combined with reference-image workflows to lock Kira's identity before a single frame is animated.

Step 0: Define Kira — The Character Spec

Before touching any AI tool, we need to lock every visual detail of Kira Vasquez on paper. The more specific we are, the more consistent every model will render her.

Face & Identity

  • Latina woman, early 30s
  • Strong jawline, high cheekbones
  • Dark brown eyes, intense gaze
  • Dark hair pulled back tight (race braid or bun under helmet)
  • Minimal makeup — this is a racer, not a model
  • Small scar above left eyebrow (earned character detail)
  • Expression range: focused calm → defiant resolve

Race Suit & Helmet

  • Matte black fire suit with subtle charcoal panels
  • No sponsor logos (she races independently)
  • Single accent: thin cyan stripe down left arm (her color)
  • Helmet: matte black with dark visor, no decals
  • Visor: tinted dark, reflective — we see lights in it
  • Gloves: black leather, worn at the fingertips

The Car

  • Matte black with carbon fiber accents
  • Low, aggressive, 2035-era prototype GT shape
  • AEGIS module visible as a cyan-lit panel on the dash
  • No external LED decorations (unlike AI-assisted rivals)
  • Car number: 12 (her number)
  • When AEGIS is killed: all cyan lights go dark

Visual Anchors to Lock

These details must remain identical in EVERY shot — they're what tells the audience "this is the same person":

  • The visor reflection pattern (neon lights in dark glass)
  • The cyan arm stripe (her one signature color)
  • The MANUAL toggle switch (physical, analog, red guard)
  • Her eye color and skin tone (the close-ups)
  • The helmet shape (smooth, no vents, matte)
The 5-Step Character Lock Pipeline

This is the exact workflow to establish and maintain Kira's identity across every model and every shot. Think of it like casting an actor, then making sure wardrobe and makeup match on every shoot day.

🎨
Generate Hero Stills
Soul / NBP / Seedream
🔒
Train Soul ID
Higgsfield Soul ID
🎬
Build Storyboard
Popcorn
▶️
Animate w/ Refs
Veo / Sora / Kling / etc.
Enforce in Post
Kling O1 + Face Swap
1
Generate the Hero Reference Sheet
SOUL / NANO BANANA PRO / SEEDREAM 4.5

Before any animation, create 10–20 static reference images of Kira from multiple angles and lighting conditions. This becomes the "casting portfolio" that every other tool references.

What to generate:

— Front face, neutral expression, even lighting (the "passport photo")
— 3/4 view left and right
— Profile view
— With helmet on, visor up
— With helmet on, visor down (reflections visible)
— Full body in fire suit, standing
— Cockpit seated position, hands on wheel
— Close-up of hands in gloves on steering wheel
— The car exterior from 3 angles

Which model to use for each:

Nano Banana Pro for face + helmet shots (4K, reasoning-guided, character consistency built in, can render the visor HUD text)
Soul for fashion-grade portraits (cinematic lighting, editorial quality, the "hero beauty" shots)
Seedream 4.5 for the car and full-body shots (consistent lighting, identity preservation, strong spatial understanding)

💡
Generate all reference images in the same session with the same prompt structure. Use the same lighting description ("dramatic side light from left, dark moody background, night rain") across ALL prompts. Consistency in prompting = consistency in output.
[LOCKED IDENTITY BLOCK — use in every prompt] Latina woman race car driver, early 30s, strong jawline, high cheekbones, dark brown eyes, dark hair pulled back tight, small scar above left eyebrow, matte black fire suit with thin cyan stripe on left arm, {ANGLE}, {EXPRESSION}, dramatic side lighting from left, dark moody background, night rain atmosphere, photorealistic, 2035 near-future motorsport, anamorphic
2
Train the Soul ID
HIGGSFIELD SOUL ID

Take your best 10–15 reference images from Step 1 and train a Soul ID avatar. This creates a digital identity that Higgsfield's system "remembers" — it preserves facial structure, bone structure, skin tone, and expression style across all future generations.

Training requirements:

— Upload 10–15 of your best Kira images
— Include multiple angles (front, 3/4, profile)
— Include multiple lighting conditions
— Include both with-helmet and without-helmet shots
— Avoid pure white backgrounds (hurts depth/blending)
— Training takes ~5 minutes, costs ~$3

After training, validate: Generate a 4–8 image grid changing ONLY the camera angle. Confirm face, bone structure, eye color, and scar are stable across all shots. If anything drifts, retrain with better reference images.

🔑
Soul ID is the single most important step. Once trained, this avatar becomes the "source of truth" for Kira across the entire film. Every subsequent tool references this identity. Train once, use everywhere.
3
Build the Storyboard with Popcorn
HIGGSFIELD POPCORN

Use Popcorn to generate all 14 key frames as a connected storyboard sequence. Popcorn's "intelligent scene continuity" system is specifically designed to maintain character identity, lighting coherence, and style across multi-frame sequences.

How to use it:

— Feed your Soul ID avatar as the character reference
— Use Auto Mode: write a detailed prompt describing the scene, specify up to 8 frames
— Popcorn's Character Anchoring memorizes facial structure, clothing texture, posture
Lighting Continuity system locks the mood once established (night rain, neon)
Style Coherence prevents the "patchwork" look between frames

Generate storyboard frames for all 14 shots. These become the input key frames for video animation in Step 4.

🎬
Popcorn + Soul ID is Higgsfield's native answer to the character consistency problem. Use Popcorn for ALL your storyboard frames — don't generate key frames independently in different models. Let Popcorn be the single source of connected visuals.
4
Animate with Image-to-Video + References
VEO 3.1 / SORA 2 / KLING 3.0 / SEEDANCE 2.0 / HAILUO 02 / WAN 2.5

Now take each Popcorn storyboard frame and animate it using image-to-video mode in the appropriate model. The key: you're not generating from text alone — you're feeding the locked key frame as the starting image.

Critical rules for consistency during animation:

Always use image-to-video, never pure text-to-video (text-to-video ignores your character lock)
— Feed the Popcorn storyboard frame as the input image
— For models that support it, also upload the Soul ID reference images as additional character refs
— Use start/end frame control (Kling 3.0, Kling O1) to define exact transitions
— Keep prompts focused on motion and camera, not character description (the image already defines the character)
— Minimize re-describing Kira's face in text — let the reference image do the work

Which model gets which reference method?
Veo 3.1
Single input image (Popcorn frame). Strong at preserving the source. Add motion + environment in prompt.
Sora 2
Single input image. Best for tight close-ups where face must stay locked. Its temporal consistency preserves identity over motion.
Kling 3.0
Start frame + end frame. Up to 7 reference images. Use Soul ID refs in the Element Library for character lock.
Kling O1
Element Library — upload all Soul ID refs. O1 "remembers" Kira across every shot. Best cross-shot consistency. Also: semantic editing to fix inconsistencies in existing clips.
Seedance 2.0
Multi-reference input (images + video). Feed Kira refs + car refs for full scene consistency. Best for montage sequences.
Hailuo 02
Single input image. Strong instruction following — describe motion precisely. Physics will be accurate; face may need post-correction.
Wan 2.5
Image-to-video + audio prompt. Upload Popcorn frame, describe dialogue/sound. Lip-sync will follow the face in the image.
Nano Banana Pro
Multi-image fusion (up to 8 refs). Generate new key frames with Kira's face locked. Best for HUD shots that need rendered text + consistent character.
5
Enforce Consistency in Post-Production
KLING O1 / CHARACTER SWAP / FACE SWAP / POPCORN

Even with perfect references, some shots will drift. That's normal. Here's the safety net:

Kling O1 Semantic Editing: If a generated clip has the right motion but Kira's face shifted, use O1's edit mode to fix it. Upload the clip + Soul ID reference → "Replace the character's face with the reference, keep everything else unchanged." O1 does this without masking or rotoscoping.

Character Swap 2.0 / Face Swap: Higgsfield's dedicated face replacement tools. Feed a generated video clip + Soul ID reference → the tool replaces the face while preserving lighting, emotion, and motion. Use this as the last line of defense for any shot where the face doesn't match.

Popcorn Re-generation: If a key frame is off, don't start from scratch — regenerate just that frame in Popcorn with the same Soul ID and scene continuity. Popcorn's anchoring system will pull it back into alignment.

The rule: If a shot's motion and atmosphere are perfect but the face is 80% right, fix the face in post rather than re-generating the entire shot. This saves massive time.

Budget for a "consistency pass" — review all 14 shots side-by-side before final edit. Flag any where Kira looks different. Fix those with O1/Face Swap. This pass is non-negotiable.
Model Consistency Matrix

How each model handles character identity — and what reference method to use.

Model Native Consistency Reference Method Max Refs Fix Strategy
Soul ID ★★★★★ Trained avatar 10–15 training photos 15+ Source of truth — retrain if drift
Popcorn ★★★★★ Multi-frame anchoring Soul ID + prompt 8 frames Regenerate single frame
Kling O1 ★★★★☆ Element Library Upload refs to Element Library 7 images Semantic edit to fix face
Kling 3.0 ★★★★☆ Subject reference Start/end frame + image refs 7 images Re-gen with locked start frame
Nano Banana Pro ★★★★☆ Reasoning-guided Multi-image fusion 8 images Re-gen with more refs
Seedance 2.0 ★★★☆☆ Multimodal refs Image + video refs Multiple Feed more refs, re-gen
Sora 2 ★★★☆☆ Temporal consistency Single input image 1 Face Swap in post
Veo 3.1 ★★★☆☆ Good preservation Single input image 1 Face Swap in post
Hailuo 02 ★★☆☆☆ Physics > faces Single input image 1 Face Swap critical
Wan 2.5 ★★★☆☆ Good with ref Image-to-video + prompt 1 Face Swap if needed
Seedream 4.5 ★★★★☆ Identity preservation Multi-image editing Multiple Re-gen with refs
The 7 Golden Rules of Character Consistency

1. Never Use Pure Text-to-Video

Always feed an input image. Text-to-video generates a new face every time. Image-to-video preserves the face you've locked.

2. One Face, Many Angles — Before Anything Else

Generate the reference sheet (Step 1) before touching any video model. This is casting day. Don't skip it.

3. Train Soul ID Early, Validate Often

Train once. Then generate 8 test images at different angles. If any are off, retrain. Soul ID is your foundation.

4. Lock the Prompt Identity Block

Write one frozen character description and paste it into every single prompt. Never paraphrase — exact words = consistent output.

5. Use Popcorn for Connected Frames

Don't generate storyboard frames in 5 different models. Use Popcorn's scene continuity system to create ALL key frames as a connected sequence.

6. Kling O1 Is Your Safety Net

If a shot drifts, use O1's Element Library + semantic editing to fix the face without re-generating the motion. Save your best takes.

7. Budget a Consistency Pass

Before final edit: lay all 14 shots side-by-side. Flag any where Kira looks different. Fix with Face Swap / O1. This step is non-negotiable.

Shot-Specific Consistency Notes

Shots 01, 11 (ECU Face) — Highest risk. These are the close-ups where any face drift is immediately visible. Use Soul ID + Sora 2 (best temporal face consistency) or Kling 3.0 (lip-sync). Always validate against reference sheet.

Shots 02, 05, 12, 13 (Wide) — Lower risk. Kira is small in frame or in a helmet. The car and environment matter more than facial detail. Focus on matching the helmet shape, arm stripe, and car design.

Shots 03, 07, 09, 10 (Cockpit POV) — Medium risk. We see gloves, hands, steering wheel, HUD — but not much face. The visual anchors here are the gloves, the MANUAL toggle, and the cyan arm stripe. Lock these props in Popcorn frames.

Shots 06 (Overtake Montage) — Quick cuts mean small inconsistencies are hidden. Use Seedance's multi-reference system to keep the car consistent. Face barely visible at speed.

Shot 08 (Near-Spin) — The car is the character here. Lock the car design using Hailuo 02 or Sora 2's physics engine. Face not visible.