FLAT SOLES SITTING POSE - ANIMA

After extensive testing with the Anima model, I found that achieving a natural sitting pose—especially one where the character’s bare soles rest completely flat against the floor—can be surprisingly difficult. The model often tilts the soles upward, bends the ankles unnaturally, or generates soles that do not properly align with the ground plane. To address this issue, I created this LoRA specifically to improve foot positioning in seated poses. It helps guide the character’s feet into a flatter and more stable position, keeping the soles naturally aligned and parallel to the ground while preserving the overall pose and anatomy of the character.For best results, I recommend using a LoRA strength between 0.7 and 1.0, depending on how strongly you want it to influence the character’s foot placement and anatomy. Although this LoRA was primarily designed to improve seated poses with the feet resting flat on the ground, you can also experiment with different prompts and body positions to see how it behaves in other poses. In many cases, it can help preserve the overall structure and completeness of the feet, maintain a clearer five-toe shape, and reduce common generation issues such as missing toes, fused toes, distorted soles, twisted ankles, or other forms of foot deformation. The final result may still vary depending on the base model, prompt wording, pose complexity, camera angle, and LoRA weight, so testing several prompt variations is recommended.To achieve the most consistent flat-footed result, I generally recommend using an image composition that is closer to a horizontal or landscape aspect ratio. A wider frame usually gives the model more room to display the full seated pose, including the legs, ankles, and feet, which makes it easier for the soles to remain fully visible, properly aligned, and naturally pressed against the ground. Vertical or portrait compositions can still work, but the tighter framing may sometimes crop the lower body or create stronger perspective distortion around the feet. You are welcome to experiment with both horizontal and vertical formats, as the final result can also vary depending on the pose, camera angle, framing, prompt wording, and resolution.