Let’s test the new model(V7):

1.LoRA Compatibility: The compatibility of character LoRA compared to V1 is generally good, but the face shape tends to be smaller than the original character (sometimes vertically smaller, sometimes horizontally narrower).

2.Thickness of Painting (Impasto Intensity): With further study of colors, I can now more clearly describe the difference between this version’s thick-painting style and the non-thick-painting versions like V1 and V2. If you add sidelighting in the prompt and have the character facing straight-on, you’ll notice that the reflections in this version are basically uniform with limited gradation, unlike real lighting which usually has more layers of gradation. At the same time, in areas not illuminated by light, the skin tone remains the same as the unlit base color, which creates a sense of heaviness. This is more of a personal preference, so it’s hard to say which is better or worse.

3.Background: When this thick-painting style is placed against a background, the light-and-shadow characteristics instead produce something like a hybrid between a heavy oil painting and a transparent watercolor painting—a result that looks bulky yet somewhat translucent. (Can anyone tell me which real-world pigment this effect corresponds to?)

For a detailed comparison, please refer to the images below.

have plan creating model furry and dragon