What problem does it solve?
This Skill helps you diagnose why a Higgsfield generation fails or produces poor output quality and guides you to the most likely prompt, model, or motion-control fixes.
Core Features & Use Cases
- Character consistency repair: resolves face drift, morphing, and identity instability by using Soul ID references and removing conflicting appearance descriptions.
- Prompt-structure correction: improves results when the prompt is ignored or doesn’t match by trimming length, removing contradictions, and enforcing a clear Subject → Action → Camera → Style ordering.
- Camera and motion control fixes: corrects generic or non-working camera movement by requiring exact preset names, one primary move, and proper handling of image-to-video animation.
- Style and VFX tuning: fixes wrong or cartoonish style by aligning model choice with preset intent and adding concrete cinematic/lighting/color-grade constraints.
- Audio and background music troubleshooting: restores lip-sync and prevents uploaded audio from being overridden by using timestamp anchoring and removing ambient SFX/music tokens.
- Safety filter mitigation: reduces blocked dark/horror generations by phrasing outcomes with atmosphere language instead of explicit acts.
Quick Start
Use the higgsfield-troubleshoot Skill when your generated video looks wrong, then regenerate after applying the specific fix that matches the observed symptom (face drift, camera not moving, prompt mismatch, style issues, or audio sync problems).
