Extending Qwen Image 2.1 Turbo's magic 8-step sigmas to 12-14 steps
GTManiK · reddit · 2026-10-11
A detailed ComfyUI workflow for Qwen Image 2.1 Turbo that extends the official 'magic' 8-step sigma curve to 10/12/14 steps:
- Problem: The official Diffusers sigmas ([1.0, 0.978453, ..., 0.0]) are fast but weak on fine details like skin and composition; RES4LYF's Custom Sigmas node breaks the curve when interpolating.
- Solution: A DIY interpolation built with rgthree's Power Puter expression node (plus RES4LYF and KJ nodes), producing a longer curve with the exact same shape—the 'Magic AI Slop Scheduler'.
- Recommended settings: CFG > 1 (3.5–5), 12–14 steps, eulerancestral (best) / euler / ersde, at 2 megapixels.
- Bonus: A 'fake SHIFT' parameter to nudge the curve—negative values (-1.0/-2.0) add fine detail, but don't deviate too far.
- Workflows are published on Civitai as 2MP images you can drag straight into ComfyUI, with full explanations inside.
More from Multimodal
- Building a physically realistic grand piano simulator down to hammer felt hysteresis with Grok — doodlestein · 2026-10-11
- AI Trailer Wins Inaugural $2.5M Future Vision XPRIZE, Heading for $10M Feature Development — lmoroney · 2026-10-11
- Open-Source Afflatus Turns ComfyUI Workflows into One-Click Forms for Image and Video Generation — Basic_Aside_7547 · 2026-10-11
- Seedance 2 Video Demo: The Missing Piece for Your Computer — Ok-Nerve941 · 2026-10-11
- ComfyUI on a $899 M4 Mac Mini 16GB: Full Benchmark Results and Workflows — FaatmanSlim · 2026-10-11
- Prime Video picks up AI-generated spin-off of Turkish hit Çukur, every frame AI-made — lmoroney · 2026-10-11