HEIF
The JPEG format has reigned over the digital imaging world since 1992. Mobile devices, computers, servers—all use this format that has established itself as a de facto standard. But this hegemony conceals technical weaknesses that have been revealed over time: no support for high color depth, no high dynamic range management, no transparency. And then there are those characteristic visual artifacts, that compression which shows its limits in the face of current needs.
In 2013, the technical team at Nokia Technologies began considering an alternative. They defined the requirements, identified use cases. The specification development took a year and a half, until the summer of 2015. But the project’s fate shifted with Microsoft’s acquisition of Nokia. The company transferred its rights and information to Apple, which took over.
HEIF (High Efficiency Image File Format) breaks with traditional image format design. Where JPEG stores a single image, HEIF can contain multiple images, entire sequences, and their metadata. The structure relies on HEVC (H.265) video coding technologies, whose advanced compression techniques are applied to still images. It’s a paradigm shift.
In 2017, Apple took the plunge. The iPhone 7, the sixth generation iPad, Macs running OS X High Sierra adopted the format. The company introduced the .heic extension for HEVC-encoded images, reserving .heif for files using other codecs. Millions of users discovered the format without really noticing.
The gains were immediate: a 50% reduction in file size compared to JPEG, with equivalent or superior visual quality. The format handles burst capture, simultaneous recording of videos and still images, animations, focal and exposure images in a single file. Editing operations on predefined derived images are made possible.
HEIF’s architecture inherits from the ISO/IEC 14496-12 base format. The box system enables data nesting, creating a hierarchy of relationships. The first box, named “ftyp,” contains general encoding metadata. A “meta” box holds the remaining information. The format integrates image tiling, which allows parallel decoding or partial loading of regions of interest in high-resolution images.
Apple adapted the specifications in its own way. The company systematically uses 512×512 pixel tiles, generating “infe” entries for each tile. These tiles are assembled into a single image via a “grid”-type derived image. Apple exploits auxiliary depth maps, accompanied by XMP metadata.
Format adoption is gradually expanding. Android has integrated it since version 9 (Pie), Samsung offers it on certain models like the Galaxy S10. Windows 10 introduced it in version 1809, but specific extensions must be installed from the Microsoft Store. The transition requires technical and practical adjustments.
For mobile photography, the impact is considerable. Users keep more photos on their devices without sacrificing visual quality. Non-destructive editing is more accessible, metadata management more flexible. Application developers have new possibilities at their disposal.
On the digital forensics side, typically for criminal investigations, HEIF complicates the work. With its structure’s ability to conceal data in different boxes, to manipulate relationships between images, investigators must develop new methodologies, account for multiple data layers, variable implementations across manufacturers.
The ISO/IEC 23008-12 standard now defines the format’s technical specifications. This standardization guarantees interoperability between different implementations. Successive updates have introduced support for predictive image coding, bursts, region annotation. Development continues.
The HEIF format adapts to the performance constraints of modern devices. But the transition from JPEG is gradual. Compatibility issues slow adoption, though the technical advantages are undeniable. Change takes time, and the outcome remains uncertain.