YOGA Model Python API

This module allows to convert and optimize 3D models.

Usage

converting and optimizing a model:

import yoga.model
yoga.model.optimize("./input.fbx", "./output.glb")

You can also tune the output by passing options:

yoga.model.optimize("./input.fbx", "./output.glb", options={
    # Model options
    "output_format": "glb",                # "glb"|"gltf"
    "fallback_texture": None,              # None|<FileLike>|"./image.png"
    "no_graph_optimization": False,        # True|False
    "no_meshes_optimization": False,       # True|False
    "no_textures_optimization": False,     # True|False
    "no_fix_infacing_normals": False,      # True|False

    # Images (textures) options
    "image_output_format": "orig",         # "orig"|"auto"|"jpeg"|"png"|"webp"|"webpl"
    "image_resize": "orig",                # "orig"|[width,height]
    "image_jpeg_quality": 0.84,            # 0.00-1.0
    "image_webp_quality": 0.90,            # 0.00-1.0
    "image_opacity_threshold": 254,        # 0-255
    "image_png_slow_optimization": False,  # True|False
})

Available Model Options

output_format

The format of the output model.

yoga.model.optimize("./input.fbx", "./output.glb", options={
    "output_format": "glb",
})

The following formats are supported:

glb (default)
gltf

fallback_texture

This option allows you to provide a fallback texture that will be used when YOGA is unable to find one of the model textures.

The following values are allowed:

None: no fallback texture,
a Python str (and unicode for Python 2): path of a fallback image file (e.g. "./fallback.png"),
a “File-like” object (file, BytesIO, etc.).

yoga.model.optimize("./input.fbx", "./output.glb", options={
    "fallback_texture": None,
})

yoga.model.optimize("./input.fbx", "./output.glb", options={
    "fallback_texture": "./fallback.png",
})

image_file = open("./fallback.png", "rb")
yoga.model.optimize("./input.fbx", "./output.glb", options={
    "fallback_texture": image_file,
})

no_graph_optimization

Disables empty graph nodes merging.

yoga.model.optimize("./input.fbx", "./output.glb", options={
    "no_graph_optimization": False,
})

no_meshes_optimization

Disable mesh optimization.

yoga.model.optimize("./input.fbx", "./output.glb", options={
    "no_meshes_optimization": False,
})

no_textures_optimization

Disable texture optimizations (textures will not be optimized using YOGA Image).

yoga.model.optimize("./input.fbx", "./output.glb", options={
    "no_textures_optimization": False,
})

no_fix_infacing_normals

Disables the assimp’s infacing normals fix. This postprocess tries to determine which meshes have normal vectors that are facing inwards and inverts them. See the assimp documentation for more details.

yoga.model.optimize("./input.fbx", "./output.glb", options={
    "no_fix_infacing_normals": False,
})

Available Image Options

YOGA Model optimizes textures using YOGA Images, so there are options equivalent to the YOGA Image ones available:

The YOGA Model image_output_format option is equivalent to the output_format of YOGA Image,
The YOGA Model image_resize option is equivalent to the resize of YOGA Image,
The YOGA Model image_jpeg_quality option is equivalent to the jpeg_quality of YOGA Image,
The YOGA Model image_webp_quality option is equivalent to the webp_quality of YOGA Image,
The YOGA Model image_opacity_threshold option is equivalent to the opacity_threshold of YOGA Image,
The YOGA Model image_png_slow_optimization option is equivalent to the png_slow_optimization of YOGA Image,
The YOGA Model image_enable_quantization option is equivalent to the enable_quantization of YOGA Image,
The YOGA Model image_quantization_dithering_level option is equivalent to the quantization_dithering_level of YOGA Image,
The YOGA Model image_quantization_max_colors option is equivalent to the quantization_max_colors of YOGA Image,

See YOGA Image options for more information.

API

yoga.model.optimize(input_file, output_file, options={}, textures=None, verbose=False, quiet=False)

Optimize given model.

Parameters:

input_file (str,file-like) – The input model file.
output_file (str,file-like) – The output model file.
options (dict) – Optimization options (see above).
textures (dict) – A dictionary that maps textures path to bytes. When not None, there will be no file system reads in order to find referenced textures. YOGA will look into that dictionary instead.
verbose (bool) – If True, Assmimp debug message will be print to stdout.
quiet (bool) – If True, YOGA will not write any warning on stdout or stderr.

Example textures dictionary:

textures = {
    "images/texture1.png": open("./images/texture1.png", "rb").read(),
    "texture2.png": open("./texture1.png", "rb").read(),
}