Search Results
7/1/2025, 11:12:06 AM
>>105763911
like this but it's always WIP. between 0.3 and 0.4 denoise, steps 10-14, bit of noise manipulation via detail daemon, either x1.5 or x2 size. some flux loras might work ok for the initial gens but I found that removing them or lowering the strength for the upscale seemed to yield better results.
sampler/scheduler, I need to run various XY plots to figure that out - not done it yet.
like this but it's always WIP. between 0.3 and 0.4 denoise, steps 10-14, bit of noise manipulation via detail daemon, either x1.5 or x2 size. some flux loras might work ok for the initial gens but I found that removing them or lowering the strength for the upscale seemed to yield better results.
sampler/scheduler, I need to run various XY plots to figure that out - not done it yet.
Page 1