ComfyUI Extension: ComfyUI-Painter-I2V-AIO

Authored by LDNKS094

Created

Updated

0 stars

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

All-in-One collection of Painter's ComfyUI nodes for video generation

Looking for a different extension?

Custom Nodes (5)

README

English | 中文

ComfyUI-PainterAIO

All-in-One collection of Painter's ComfyUI nodes for Wan 2.1/2.2 video generation.

This project consolidates and refactors multiple ComfyUI video node projects with unified API and optimized code structure.

Based On

| Project | Author | Feature | |---------|--------|---------| | ComfyUI-PainterI2V | princepainter | I2V with motion enhancement | | ComfyUI-PainterI2Vadvanced | princepainter | Dual sampler + color protection | | ComfyUI-PainterLongVideo | princepainter | Long video extension | | Comfyui-PainterFLF2V | princepainter | First-last frame interpolation | | ComfyUI-Wan22FMLF | wallen0322 | Multi-frame reference conditioning |

Nodes

| Node | Description | |------|-------------| | PainterI2V | Image-to-video with I2V / FLF2V modes, motion enhancement | | PainterI2VAdvanced | Dual-phase sampler (high/low noise separation), supports loop continuation | | PainterI2VExtend | Video extension for long video generation | | PainterSampler | Sampler with motion enhancement | | PainterSamplerAdvanced | Dual-phase sampler for PainterI2VAdvanced |

Installation

cd ComfyUI/custom_nodes
git clone https://github.com/LDNKS094/ComfyUI-PainterAIO.git

Features

  • Motion Enhancement: Fix slow-motion issues with 4-step LoRAs (e.g., lightx2v)
  • Color Protection: Prevent green/dark color drift after motion enhancement
  • Dual-Phase Sampling: High noise for motion, low noise for semantic reference (Advanced only)
  • First-Last Frame Interpolation: FLF2V mode with start and end frame anchoring
  • Long Video Extension: Seamless continuation from previous video segment
  • SVI LoRA Compatible: Supports latents_mean padding for SVI mode
  • Loop Support: Auto-conversion between latent/image for ComfyUI loop workflows

Node Details

PainterI2V

Basic I2V node with motion enhancement.

| Parameter | Description | |-----------|-------------| | motion_amplitude | 4-step LoRA fix. 1.1-1.2 normal, 1.2-1.5 fast motion | | color_protect | Prevents color drift from motion enhancement | | svi_mode | SVI LoRA mode with latents_mean padding | | start_image | First frame reference | | end_image | Last frame for FLF2V mode | | clip_vision | Semantic guidance |

PainterI2VAdvanced

Advanced node with dual-phase sampling (high/low noise separation).

Outputs 4 conditionings: high_positive, high_negative, low_positive, low_negative

Use with PainterSamplerAdvanced for dual-phase sampling.

High Noise Phase (Motion)

  • motion_amplitude - Motion enhancement strength
  • color_protect - Color drift prevention
  • end_image - End frame anchoring

Low Noise Phase (Semantics)

  • reference_latent - From start_image (automatic)
  • clip_vision - Semantic guidance
  • correct_strength - Reference correction strength (0.01-0.05 recommended)

Continuation Parameters (Standard Mode Only)

  • overlap_frames - Frame overlap count (4-8 recommended)
  • continuity_strength - Motion lock strength (0.1-0.2 recommended)
  • previous_latent or previous_image - Previous segment for continuation

Mode Differences

| Feature | Standard Mode | SVI Mode | |---------|---------------|----------| | Continuation source | previous_image | previous_latent | | overlap_frames | Used (4-8) | Fixed (1 latent) | | Concat padding | Grey (0.5) encoded | latents_mean (zero) |

PainterI2VExtend

Video extension node for long video generation.

| Parameter | Description | |-----------|-------------| | overlap_frames | Frame overlap for continuity (4-8 recommended, standard mode only) | | motion_amplitude | 4-step LoRA fix (both modes) | | color_protect | Color drift prevention (both modes) | | svi_mode | SVI LoRA mode with anchor + last latent | | previous_video | Previous video segment to continue from | | anchor_image | Anchor frame (defaults to previous_video[0]) | | end_image | Target end frame | | clip_vision | Semantic guidance |

Parameter Guide

motion_amplitude

| Motion Type | Recommended Value | |-------------|-------------------| | Subtle motion (talking, blinking) | 1.0 - 1.1 | | Moderate motion (walking, gestures) | 1.1 - 1.2 | | Large motion (running, jumping) | 1.2 - 1.5 |

color_protect

Keep enabled (default: True) to prevent color drift after motion enhancement.

overlap_frames (Advanced/Extend)

Number of pixel frames to overlap for continuity. Internally converted to latent index (overlap_frames // 4).

| Use Case | Recommended Value | |----------|-------------------| | Standard continuation | 4 | | Smoother transition | 8 |

continuity_strength (Advanced Only)

Controls motion lock strength between segments.

| Use Case | Recommended Value | |----------|-------------------| | Normal continuation | 0.1 | | Stronger motion lock | 0.2 |

correct_strength (Advanced Only)

Reference latent correction strength for low noise phase.

| Use Case | Recommended Value | |----------|-------------------| | Normal | 0.01 - 0.03 | | Stronger reference | 0.03 - 0.05 |

Acknowledgements

License

MIT License

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Learn more