Common mistakes when importing normal maps for 3D models include incorrect texture channel assignments, flipped normals from coordinate system mismatches, and using low-resolution maps that lose detail.
Incorrect channels happen when the normal map is placed in a diffuse or specular slot instead of the dedicated normal channel—so the software doesn’t use it to add depth. Coordinate system differences (like OpenGL’s Y-up vs. DirectX’s Y-down) flip the map’s green channel, making surfaces look inverted. Low-res maps blur fine details, so the model’s texture depth disappears.
If your model looks flat or has weird shadows after import, first check the normal map’s channel in your software—then try flipping the green channel (most tools have a “flip Y” toggle) to fix coordinate mismatches.