What is EXR? OpenEXR High Dynamic Range Format Explained

How EXR Works

OpenEXR stores pixel values as floating point numbers rather than the fixed 0 to 255 integers used by ordinary image formats. It supports a 16-bit half float and a 32-bit full float per channel. Because floating point can represent both very small and very large magnitudes, an .exr file can describe a scene where a bright light source is thousands of times more intense than a shadowed area, all without clipping. This makes it a true high dynamic range format suitable for physically based rendering and compositing.

An EXR image is not limited to red, green, and blue. The format supports an arbitrary number of named channels, which allows a single file to carry many separate render passes from a 3D scene. Common examples include diffuse, specular, and reflection passes, a depth or Z channel that records distance from the camera, and normal or motion vector data. A compositor can then recombine or adjust these passes non-destructively, which is central to modern visual effects work.

Pixels can be laid out in scanline order for sequential access or in tiled order, where the image is divided into rectangular tiles that can be read independently. Tiled storage supports multi-resolution representations, so an application can load a lower resolution version quickly or read only the region it needs. This flexibility matters for very large frames used in film production.

OpenEXR offers several compression methods, including lossless options such as ZIP and the wavelet based PIZ scheme, and lossy options such as PXR24, which reduces float precision, and the DWA family, which targets higher compression for review and delivery. The choice lets a facility balance file size against fidelity depending on whether the image is a working master or a lighter intermediate.

History and Standardization

OpenEXR was created at Industrial Light and Magic, the visual effects studio, to meet the needs of film production where standard 8-bit formats could not preserve the range and precision required for compositing. It was released as open source in the early 2000s, which allowed the wider industry to adopt and extend it freely.

Because the source code and specification were published openly, EXR spread quickly across visual effects and animation studios and became a common interchange format between rendering and compositing tools. Over time the project added features such as multi-part files and deep data, where a single pixel can store multiple samples at different depths, which is useful for compositing semi-transparent and volumetric elements.

Stewardship of OpenEXR is now carried out under the Academy Software Foundation, an industry body that hosts several open source projects used in motion picture production. This governance helps keep the format maintained as a stable, widely trusted standard for high dynamic range imagery in film and related fields.

EXR Compared to Radiance HDR and TIFF

Compared with the Radiance HDR format, OpenEXR generally offers greater precision and far more flexibility. Radiance stores high dynamic range compactly using a shared exponent scheme, which is efficient but coarser, and it is essentially a single color image. EXR provides true per channel floating point and the multi-channel structure needed for layered render passes, which is why it has largely become the preferred format in professional visual effects.

Against TIFF, the comparison is more nuanced. TIFF is a highly flexible container that can also store floating point data, and it is common in photography and print workflows. However, TIFF lacks the standardized multi-channel, multi-pass conventions that visual effects pipelines rely on, so EXR fits that specific workflow more naturally.

The honest trade off is that EXR files, especially at full 32-bit precision with many channels, can be large and are overkill for ordinary photographs or web images where an 8-bit format is entirely sufficient. Radiance HDR remains a lightweight option for simple environment maps, and TIFF stays popular in still imaging, while EXR earns its place where wide dynamic range and layered data are essential.