HDR Formats Evaluation

First image testing software to evaluate HDR formats!

Through the HDR16 module, Analyzer offers native support for HDR input formats as well as analysis tools more adapted to these contents. These include metrics in ICtCp color space that are independent of the white point luminance value for HDR analysis and dedicated HDR to SDR tone mapping to be able to use existing SDR-oriented measures.

Color consistency, Contrast preservation and Texture evaluation in one shot
Noise and DR measurement using 120dB target
Perceptual evaluation including texture loss, moiré, aliasing, color shift and sharpness
Video Exposure Convergence
Video Color stability
Support for more than 8bit photo input (8 to 16bit HEIF, AVIF…) and video input (base layer for HLG and PQ such as HDR10, HDR10+, HDR Vivid, Dolby Vision)
Support for Gain map for HDR Photo for HEIF & JPEG files (Android Ultra HDR, Apple HDR gain map)

Recent technology improvements in both consumer and industrial cameras have enabled on-device capture and mastering of High Dynamic Range (HDR) content without user intervention. Several formats coexist for video (HDR10, HDR10+, Dolby Vision, etc.) and more recently for photo (HEIF, AVIF, JPEG-XL, PNG, with or without gain maps and/or ICC profiles, etc.), encoding an increased range of luminance (equal to or greater than 1000 cd/m2), color (through the use of wide-gamut color spaces), and contrast (equal to or greater than 1000:1) compared to the more common Standard Dynamic Range (SDR) formats. When viewed on a compatible display, these photos and videos can reproduce real-life scenes more realistically for human observers. Traditional tools for SDR analysis such as CIELAB computation are no longer sufficient to assess image quality in HDR.

New measurements & ICtCp Color Space

In today’s applications, CIE-L*a*b* is the most commonly used color space for exposure and color analysis. While this space has proved its usefulness time and time again in many contexts, it makes some assumptions on the nature of the scene. Namely, it assumes that the scene only contains reflective surfaces and that is it lit by a single illuminant of known color. Unfortunately, these assumptions do not always hold true, particularly in the case of HDR scenes, where different light sources of different color need to be captured at once.

To handle these situations, Analyzer now fully supports the more modern ICtCp color space. This color space does not suffer from the same limitations and allows for more precise analysis of HDR content.

Analyzer v9 supports all most common HDR image formats and can compute color measurements in the ICtCp color space, particularly suitable for HDR content. These results include the familiar displacement graphs (geometric distance of the measured points from their reference) in this new space, but also an estimation of the device transfer function (i.e., the luminance on screen as a function of the luminance in the scene).