Model import settings

Author: Charley && Meng Xingyu

1. Model import and setting

LayaAir3-IDE only supports the import of model formats with suffixes such as fbx, gltf, and glb.

LayaAirIDE also supports the import of models and meshes exported by the "LayaAir-Unity Resource Export Plug-in" (.lh and .lm suffixes). This article mainly refers to external model files. If there is a need for lm grid configuration, this document is also applicable.

Drag the models with these suffixes into the assets directory of the project resources, and they will be automatically converted and imported, and then they can be added to the scene.

The operation is as shown in the animation 1-1:

(Animation 1-1)

When a model is selected, you can modify the settings for model import in the Property Settings Panel. Click Apply after each modification, and the import will be re-imported based on the modified properties.

The operation is shown in Figure 1-2:

(Figure 1-2)

2. Model scene (Scene) import settings instructions

A model may contain a lot of information, such as scene-related information such as cameras and light sources, as well as meshes, animations, materials, bones, etc. We classify this information, starting with the scene-related settings.

2.1 Model scaling scale Factor

Model scaling is mainly used to preprocess the model size according to the target scene scale to make it meet the design requirements and optimize the scaling relationship between different models.

For example, you can adjust the model to the physical size that meets your needs through the settings here. The effect comparison is shown in Figure 2-1:

(Figure 2-1)

2.2 Model unit conversionconvert Units

The default unit after the model is imported is meters.

If you change the unit to centimeters, there will be a checked associated option Size Unit Converter that is checked by default. The purpose of the default check is to maintain the visual size of the current model and only change the unit of the model.

If checked and unchecked, the scene will feel magnified 100 times. The effect comparison is shown in Figure 2-2:

(Animation 2-2)

2.3 Import model built-in data (visibility, camera, light source)

If the art has visibility, camera, and light attribute settings when making the model, LayaAir3-IDE will not import them by default. It is recommended to add and set them in the scene of LayaAir-IDE.

When developers have these built-in data requirements for using the model, they can also check these options, as shown in Figure 2-3. Re-import it by clicking Apply.

(Figure 2-3)

2.4 Add level of detail add Lod Group

When Lod data exists in the model, checking this option will export the Lod data and automatically create a Lod Group component to set the Lod data when using the model. The effect after adding is shown in the animation 2-4.

(Animation 2-4)

3. Instructions for importing the model’s mesh (Mesh)

3.1 Easier-to-understand attribute descriptions

There are some attribute settings that are easier to understand, as shown in Figure 3-1.

(Figure 3-1)

We provide an overview via tables.

3.2 Index data format Index Format

Index data is used to determine the position of vertices in the vertex buffer, thereby building the shape of the geometry. This property is used to control the precision and range of these index data, thereby affecting the performance and quality of rendering.

As shown in Figure 3-2, this attribute has three options: Auto, UInt16-bit, and UInt32-bit.

(Figure 3-2)

Auto: Automatic is the default option. In this mode, LayaAir will automatically decide which data format to use based on the number of vertices. By default, UInt16 will be used first. When the number of vertices exceeds 65535, it will automatically switch to UInt32 data format. .

UInt16: This means that the index data is represented by a 16-bit unsigned integer (0~65535). Using a 16-bit index saves memory and is sufficient to express the vertex index in most cases. However, when the number of vertices of the model exceeds 65535, it needs to be split into multiple sub-meshes for rendering, which may increase the number of draw calls.

UInt32: This means that the index data is represented using a 32-bit unsigned integer (0~4294967295). Using 32-bit indexes can support larger models and avoid splitting into multiple subgrids, but will take up more memory.

3.3 Processing normals and tangents

In terms of the ability to handle normal effects, there are mainly three related attributes, namely Normal, Normal Calculation Mode, and Tangent, as shown in Figure 3-3:

(Figure 3-3)

3.3.1 Normal

The Normal attribute has three options, importimport, calculatecalculate, and NoneNone. As shown in Figure 3-4.

(Figure 3-4)

Import means using the predefined normal data in the model file. This means that if the model file contains normal information, LayaAir will directly use these normal data for rendering. If the model has no normal data or the normal data is incorrect, you may get incorrect lighting effects.

Calculation means that the normal data of the model will be recalculated when the model is imported. This may be useful in certain situations, such as when the model's normal data is incomplete or inaccurate. Recalculating normals may consume some computing resources, but it ensures that the model is lit correctly.

Whether it is importing or calculating, you need to set the associated attribute Normal Calculation Mode, and selecting None will completely ignore the normal data of the model, and the vertices will not be imported. The attribute Normal Calculation Mode will not be displayed in the IDE. None this option may be used for specific rendering needs, either by manually editing the normals in LayaAir or using a script to calculate the normals. Failure to do so means that when rendering, no normal data will be available, potentially resulting in incorrect lighting visuals.

3.3.2 Normal Calculate Mode

Normal Calculation Mode has two options: Unweight and Area Weight. As shown in Figure 3-5.

(Figure 3-5)

Unweight, meaning the normal will be calculated as a simple average of adjacent vertices. This means that the normal of each vertex is only affected by its neighboring vertices, regardless of the size or angle of the face. This works well for smooth surfaces, but normal discontinuities may occur at sharp corners or separated edges.

Area Weight will weight the average of the normals of adjacent vertices according to the size of the face. This produces smoother normal transitions on sharply angled surfaces, reducing normal discontinuities.

3.3.3 Tangents

Tangent data is a key part of advanced effects like normal maps and normal mapping. Normal maps simulate high detail by storing normal information on the model's surface. Tangent data helps calculate pixel normals, giving objects visual bump detail. So Tangent is the same as Normal Calculation Mode,only when the normal line exists (not set to None), it will be displayed on the property panel.

Tangents define the tangent vectors for each vertex on the model surface. They are None, calculate, and import respectively. As shown in Figure 3-6.

(Figure 3-6)

None means tangent data will not be calculated or imported. This means that no tangent data will be available during rendering, potentially affecting the correctness of effects such as normal maps. If this is not a problem, select this option to save memory used by vertices.

calculate means recalculating the tangent vector when importing the model. This can be useful when using effects such as normal mapping and normal mapping, as tangent vectors need to be used with normals and texture coordinates. Recalculating tangents may consume some computing resources, but it ensures the correctness of the rendering effect.

import will use the predefined tangent data in the model file. This means that if the model file contains tangent information, LayaAir will directly use these tangent data for rendering. If the model has no tangent data or the tangent data is incorrect, you may get incorrect normal mapping effects.

3.4 UV1

When the model is imported, the UV1 that comes with the model can be imported by default. As shown in Figure 3-7,

(Figure 3-7)

You can check Normal UV1. After checking, when the model is imported, its own UV1 will have the standard for light map baking. Otherwise, the imported UV1 by default will not start the light map baking process. If UV1 is not used for light map baking, you do not need to check this option.

Sometimes, the model only has one default UV. If you want to generate another UV1 to support the light map, as shown in Figure 3-8, you can select the generate option.

(Figure 3-8)

The following is the configuration of its subkeys:

3.4.1 maxChartArea

The maximum chart area size is a parameter in the lightmap generation process that controls the maximum chart area. A graph is the basic unit used in algorithms to organize patches. This parameter limits the total area of ​​patches contained in a single chart to control the size of the chart.

This parameter plays a role in the chart generation process and is used to control the chart generation and merging process. The algorithm checks whether the area of ​​each chart exceeds the value of maxChartArea. If it exceeds, it will limit the generation or merging of charts. If the maximum chart area size is set to 0 (the default), it means there is no area limit and the chart can be any size.

Smaller values ​​result in more graphs, each containing fewer patches. Larger values ​​may produce fewer large charts.

3.4.2 Max Boundary Length

During the lightmap generation process, the maximum boundary length is used to constrain the merging and generation of charts.

In the algorithm, it will be judged whether the boundary length after generating or merging two charts exceeds the preset maximum boundary length limit. If it exceeds, it will be restricted and the merge will not be performed or the generation will be aborted to control the boundaries of the generated chart.

By default, the value of the maximum border length is 0, which means that there is no limit on the maximum border length of the chart.

3.4.3 Normal Deviation Weight

The normal deviation weight is an important parameter used to control the cost calculation in the patch charting process during the light map generation process.

During the graphing process of patches, the algorithm will traverse each face in the current graph area and determine which faces can be added as candidate faces to the graph generation. The basis for calculation is called cost.

The angular deviation between the surface normal and the best-fitting normal of the chart (the average normal vector of the entire chart) is one of the factors that affects the cost. When it exceeds a certain angle range (about 45 degrees), it is considered to be larger, is directly excluded from the generation queue. If the angle is not exceeded, the cost will be multiplied by the normal deviation weight value and used as a basis for further calculations. Therefore, the normal deviation weight is one of the important parameters that affects the normal angle deviation cost and chart generation.

When the normal deviation weight value is larger, the normal deviation has a greater impact on chart generation. A larger value makes the graph generation algorithm more inclined to select patches with more consistent normals to be assigned to the same graph, so that the generated graph is smoother in the normal direction. However, too high a value may result in chart generation that is too conservative and fails to take full advantage of favorable polygon structures.

On the contrary, the smaller the normal deviation weight value, the graph generation algorithm will tolerate a certain degree of normal deviation during the patch graphing process. This means that the algorithm may assign patches with widely different normals into the same graph, even though their normal directions are not exactly the same. Might be suitable for some scenarios that require more variation.

3.4.4 Maximum Iteration

Maximum Iteration is a parameter that controls the number of iterations during chart generation and optimization. Iteration is an important step in the chart generation algorithm, which continuously optimizes the generated charts through multiple iterations to achieve better results.

More iterations mean the algorithm will have more opportunities to adjust and optimize the chart, potentially producing a better chart layout. However, too many iterations may increase the computational cost and time overhead of the algorithm.

A lower number of iterations may be more time efficient, but the resulting graph may not be as optimized as with more iterations.

Developers can set here to control the number of iterations of chart generation and optimization to balance computational efficiency and the quality of generated results.

3.4.5 Texels Per Unit

The Texels Per Unit parameter defines the number of texels contained in each unit of space (for example, a 1x1 unit rectangle). This parameter is used during lightmap generation to control the texture density of the chart.

Specifically, it affects the size distribution of charts within texture atlases. In the lightmap generation algorithm, the value of texels per unit is taken into account when determining the position and size of the map within the texture atlas. If the value is greater than 0, the algorithm will allocate the texture area of ​​the chart according to the given value so that it has a higher resolution on the texture. If the value is 0, the algorithm estimates an appropriate value based on the surface area of ​​the chart and the given resolution so that the chart is properly represented on the texture.

In effect, this parameter affects the mapping quality of the chart in the final texture atlas. Higher values ​​will result in the chart having a higher resolution on the texture, making details clearer, but will also increase the size of the final texture atlas. Lower values ​​make the chart less resolution on the texture, but can reduce the size of the texture atlas. Therefore, adjusting the Texels Per Unit parameter can provide a trade-off between chart quality and texture atlas size.

3.4.6 Padding

Margin is the extra pixel space added around the chart. These extra pixels are used to extend the boundaries of the chart to ensure there is enough space between them and no overlap occurs.

Expanding margins provides better texture sampling boundaries, reduces color mixing between charts, and improves edge rendering.

When used, the margin values ​​can be adjusted as needed to trade off rendering quality and atlas size. Reasonable values ​​will generally provide better rendering, but increasing the margins too much may cause the atlas to become too large, taking up more memory.

4. Instructions for importing the model’s bone binding (Rig)

The bone binding when importing the model has two setting options: maximum number of bones per vertex and maximum bone weight, as shown in Figure 4-1.

(Pic 4-1)

4.1 Max Bones/Vertex

As the name suggests, this is used to set the maximum number of bones that can be bound to each vertex, that is, the maximum number of bones that each vertex can be affected by.

This parameter is directly related to the complexity and expressiveness of the character model's animation. If the value of Max Bones/Vertex is set high, it means that each vertex can be affected by more bones, and more complex and detailed animation effects can be produced. But it will also increase the computational burden of the game, which may lead to performance degradation. We need to weigh it according to the actual situation and needs.

4.2 Max Bone Weight

Maximum bone weight is used to limit the maximum weight value given by a single bone to each vertex.

What are bone weights? When we skin a character model, we need to decide how each bone affects the vertices of the model. This is the weight, for example a weight of 0.5 here means that this bone will have 50% influence on this point.

Typically developers set different weights for each bone to fine-tune how they affect the animation and shape of the model.

The Max Bone Weight here is used to limit the set bone weight to not exceed the value here. When it exceeds the value here, the weight setting here will be adopted.

5. Model animation (Animator) import settings instructions

After checking Anim Compression, as shown in Figure 5-1. The volume of animation data can be compressed according to configuration.

(Figure 5-1)

The minimum ignore value means that if the value exceeds the set value, the keyframe will be ignored by the algorithm, thereby achieving the purpose of reducing animation data.

There are three ignored dimensions, namely displacement, scaling and rotation.

Ignored input values ​​are in degrees.

6. Instructions for importing model materials (Materials)

Material position is used to set the access position source of the model material. There are two options. As shown in Figure 6-1.

(Figure 6-1)

Select Use embedded material to directly use the material embedded in the model file.

Select Use external materials and click Apply, the materials inside the model file will be extracted to the outside of the model and used as material resources outside the model.