comfyui picture blur
when i use v2.0,and my workflow is as follows, the generated pictures are blur, can you teach me how i could do,thanks very much
Recent tests show the model likely lost its speed due to long training and now needs more steps for good results. I'm exploring if distillation can restore zimage's fast inference.
I have set the steps as 20 or more,but the result picture didn't change better.
i want generate this kind of picture,however the outcome is as follows：
the input image is:
May I ask what the control image is for this?
i want generate the image like this
The circled areas indicate a sense of depth，and the lines are not purely black, and their thickness varies
my output is like this:
my prompt like this:
positive:Spatial three-dimensionality with strong light and shadow. Lighting creates depth; high contrast between light and dark gives objects volume. Light, shadow, and coloring define material textures and spatial 3D effects. Boundary line weights (thickness) must reflect three-dimensionality. Outlines should vary in thickness and shade (intensity), avoiding solid black, to enhance the 3D structure of the objects.
negtive:3D render, hard edges, realistic, metallic texture, keep original shape
it seems like cannot understand the prompt about the lines and depth