Ref2V Workflow: Mixing Match and Max Modes for Control
Formal_Courage2711 · reddit · 2026-08-25
A user discusses whether Stable Diffusion's Ref2V plugin supports mixing 'Match' and 'Max' reference modes. The trade-off is that setting everything to Max is too slow, while Match lacks detail precision for elements like faces. The goal is to apply different weights to different reference images (e.g., Max for faces, Match for backgrounds) to balance quality and performance.
More from Apps
- Anthropic updates Claude Slack agent to read full history and intervene unprompted — MichaelFNunez · 2026-08-25
- Cohere becomes default local model for SuperWhisper, enabling offline transcription — cohere · 2026-08-25
- Building a Multi-Agent Personal Assistant Workflow with Claude — michael_k18 · 2026-08-25
- Grok Voice Now Available via API and Agent Builder — SpaceXAI · 2026-08-25
- Grok Voice Tops Speech-to-Speech Index, Handling 15k Daily Starlink Calls — SpaceXAI · 2026-08-25
- Open source Midjourney Mosaic recreates classic iTunes art screensaver — zats · 2026-08-25