LabTexture

A minimal image loading and saving library with a single C header and a LabTexture struct that supports 8-bit, 16-bit float, and 32-bit float pixel formats.

Supported Formats

Reading

All formats loaded via stb_image are decoded as 8-bit RGBA. EXR, AVIF, and PFM preserve their native floating-point data.

Writing

Pixel format conversion on write:

• EXR passes the source format through natively (U8→UINT, F16→HALF, F32→FLOAT) with no allocation.
• HDR converts U8/F16 to F32 in a temporary buffer.
• PNG / JPEG / TGA / BMP convert F16/F32 to U8 (clamped 0–255) in a temporary buffer.

Usage

#include "LabTexture/lab_texture.h"

// Load an image
LabTexture texture;
LabTextureResult result = lt_LoadImage("texture.png", &texture);
if (result != LabTextureOk) {
 // handle result codes here
 // ...
}
// Use the texture data (e.g., texture.pixels_i8, texture.width, etc.)
// ...

// Save to a different format
result = lt_SaveImage("texture.exr", &texture);
// Release allocated memory
lt_ReleaseImage(&texture);

Passing NULL as the LabTexture* argument to lt_LoadImage tests whether the file is loadable at the same cost as actually loading it.

Result Codes

Build & Install

cmake -B build
cmake --build build
cmake --install build --prefix /usr/local

Consumers use it via find_package:

find_package(LabTexture REQUIRED)target_link_libraries(my_app LabTexture::LabTexture)

LICENSE

The OpenEXR code:

// SPDX-License-Identifier: BSD-3-Clause

The AVIF library:

// SPDX-License-Identifier: BSD-2-Clause

stb_image / stb_image_write:

/* public domain — no warranty implied; use at your own risk */

All other code in this library not otherwise marked is

// Copyright 2026 Nick Porcino
// SPDX-License-Identifier: BSD-2-Clause