ArtificialGuyBR

Home / Blog / Training LoRAs

Double Exposure LoRA for Flux Klein

How to use the Double Exposure REDMOND LoRA with Flux Klein to generate double-exposure composites in ComfyUI.

2 sources cited Training LoRAs

On this page

  1. What the model does
  2. How to use it
  3. Common errors and fixes
  4. FAQ
  5. Sources

Double Exposure LoRA for Flux Klein

A LoRA that adds the classic double-exposure effect to FLUX.2 Klein 9B by filling silhouette shapes with landscape, botanical, or sky imagery.

What the model does

The LoRA generates images where a dark silhouette—typically a human profile, figure with instrument, or bust—is overlaid with a second photographic layer. The result keeps the outer shape solid while the interior shows a sunset, field, forest, or floral scene. The style is cohesive across samples: high contrast, clean silhouette edges, and intentional color grading in reds, oranges, blues, or warm neutrals.

How to use it

Load the LoRA in ComfyUI with FLUX.2 Klein 9B as the base model. Include Double Exposure, DoubleExposureREDAF. in the prompt to activate the style. Adjust CFG and steps as needed for clean silhouette boundaries; the author does not publish exact numeric recommendations beyond standard Flux settings.

Common errors and fixes

Error | Cause | Fix | Source

|---|---|---|---

Style does not appear | Missing trigger words | Include Double Exposure and DoubleExposureREDAF. | HF README Weak silhouette edges | CFG too low or short decode | Increase CFG slightly; use standard Flux step count | Observation Wrong base model | Loaded on SD 1.5/XL | Use black-forest-labs/FLUX.2-klein-9B | HF tags

FAQ

Does this work in Automatic1111? The author recommends ComfyUI, but diffusers-based workflows can load the Safetensors LoRA.

Is the style consistent across subjects? Yes; samples show consistent double-exposure treatment across portraits, musicians, and botanicals.

Can I use it commercially? Yes, under the Apache 2.0 license.

Sources