From the moment an image is captured to the moment it is displayed on screen, image creation passes through multiple stages starting with the scene itself, image capture, sensor response, image processing, and final display. Each stage of this pipeline can influence how users perceive realism and form their image preferences. So, we decided to conduct a study aiming to comprehensively analyse real-world user visual experience and better understand the key question: 

“Is what users see truly what they get?” 

 

The results of our study sparked significant interest t across the smartphone industry and fuelled discussions during several expert round tables. A special focus was placed on exploring improvement potentials together with partners such as ams OSRAM, a recognized leader in advanced light sensing technologies supporting enhanced color consistency.  

“This study provides ams OSRAM with deeper insights into the key challenges OEMs face in complex imaging scenarios. Findings from DXOMARK’s testing framework,  particularly in low-light and zoom conditions, will directly inform sensor design, helping smartphones achieve more consistent and accurate color performance across varied lighting environments.” 

— ams OSRAM, Head of Research & Development Optical Sensors, Dalibor Stojkovic 

About the Glass-to-Glass Study

DXOMARK designed and conducted a series of experiments to track the full imaging pipeline from capture to display which focused on how faithfully images reproduce real scenes, how different stages of the imaging pipeline affect final rendering, and the respective roles played by the camera and the display. 

DXOMARK designed and conducted a series of experiments to track the full imaging pipeline from capture to display which focused on how faithfully images reproduce real scenes, how different stages of the imaging pipeline affect final rendering, and the respective roles played by the camera and the display. 

Key Findings from the Glass-to-Glass Study

Camera Performance: The Impact of the Capture Stage on Image Quality 

The study found significant differences among devices in their ability to accurately reproduce scenes during image capture, particularly in low-light conditions, where disparities became more pronounced. Common issues included unnatural white balance and skin tone rendering, uneven exposure between faces and backgrounds, and lack of HDR support, which led to noticeable quality degradation when images were viewed on HDR displays. 

Display Performance: Rendering Goes Beyond Simple Reproduction 

Two reference images were used for perceptual evaluation across different smartphone displays: one SDR image captured with a DSLR camera and one HDR image captured with a smartphone. This approach enabled analysis of rendering differences across display devices. 

Results showed clear differences in brightness tuning strategies across devices, directly affecting image clarity and readability. Some devices adopted lower overall brightness levels but enhanced key areas through local tone mapping. 

Color shifts were more pronounced than expected and were easily perceptible to the human eye. These differences are closely related to how devices adapt to ambient lighting. For example, some devices can adjust color temperature based on ambient light, resulting in a warmer overall appearance in low-light environments. 

As smartphone cameras approach professional-level performance, achieving accurate color reproduction remains a major challenge. Professional photography relies on precise white-point calibration, a standard now being introduced into mobile imaging. Through its collaboration with DXOMARK, ams OSRAM combines its expertise in high-precision ambient light sensing with an industry-leading image quality evaluation framework. This collaboration strengthens cooperation among manufacturers, academia, and technology providers, collectively advancing mobile imaging toward professional standards. 

— ams OSRAM, Head of Optical Sensor Application Engineering, Miljan Petkovic 

The Importance of Ambient Light and Adaptive Display Color Temperature 

Correlated Color Temperature technology is an example of this adaptation that senses ambient lighting conditions and adjusts screen color temperature to better match the environment and content. Even with color temperature adaptation enabled, devices differ in brightness and tone-mapping strategies. As a result, what users see is often the combined outcome of raw capture, image signal processing, and display adaptation. 

This further highlights the importance of conducting systematic evaluations across the entire imaging pipeline. 

Images are rarely viewed under ideal conditions. The environment serves as a critical link between captured content and real viewing experience. Image authenticity refers to the extent to which captured or displayed content preserves its original optical and perceptual characteristics. In laboratory settings, this can be quantified using controlled lighting, calibrated color charts, reference images,calibrated monitor and high-precision sensors and measurement equipment. Authenticity is typically assessed by measuring deviations from a ground-truth reference in areas such as tone mapping, color accuracy, texture preservation, and noise performance. These methods apply to both camera and display evaluation. This study further extends the approach to full-pipeline evaluation by comparing device-rendered images directly against real scenes. 

— Thibault Cabana, Display Quality Engineer, DXOMARK 

End-to-End User Experience: Does “More Real” Always Mean “More Preferred”? 

A perceptual study was conducted with a group of European consumers to explore how closely rendered images align with real-world perception. 

Key Findings 

The most realistic device is not always the most preferred realism matters, but it is not an absolute standard. 

Results indicate that devices ranked highest in perceived realism was the second most preferred device overall.  Devices perceived as deviating more significantly from real scenes were receiving lower overall user preference because of excessive brightness, less suitable contrast and color rendering, and highlight clipping as key reasons.  

The study suggests that while users can accept images that are somewhat optimized rather than strictly realistic, preference remains strongly correlated with perceived realism. Achieving the right balance between accuracy and visual appeal has become a central challenge in improving user experience. 

 There is no inherent contradiction here. While users value faithfulness to real scenes, realism alone does not dictate preference. In individual evaluations, among the devices considered close to reality, the one that achieved the highest overall preference is not the one that ranked closest to reality. This shows that realism plays a significant role but allows room for tuning. Further analysis suggests that the most preferred one adopts a more aggressive CCT adjustment strategy, better aligned with the warmer lighting conditions commonly found in Europe. In low-light environments, its relatively lower brightness combined with contrast optimization helps preserve shadow detail and improve overall visual comfort. 

— Thibault Cabana, Display Quality Engineer, DXOMARK 

Achieving a Better End-to-End Experience and the Value of Full-Pipeline Evaluation Combining Camera and Display 

Testing individual components in isolation is no longer sufficient to explain real user perception and preference. Coordination and tuning of components such as ambient light sensors, cameras, and displays directly influence user perception. User preferences are also strongly influenced by cultural and regional aesthetics; what appears natural in one market may not resonate in another. Localized evaluation therefore plays a critical role. 

Looking ahead, manufacturers can enhance overall experience by adopting differentiated display styles while maintaining imaging accuracy. To deliver meaningful viewing experiences, devices must account for displayed content, viewing conditions, and file formats (such as HDR or SDR) when selecting appropriate display parameters. 

DXOMARK has long adopted a user-centric approach and works closely with both camera and display teams across the smartphone industry. Our research shows that systematic joint evaluation between capture and display has long been missing, despite the fact that users ultimately view their photos on screens. Incorporating both capture and display into a unified evaluation framework is therefore essential. Beyond camera performance alone, high-quality display tuning is also a critical component of a strong camera user experience. 

— Frédéric Guichard, CEO, DXOMARK 

Establishing a Scalable Framework for Imaging Evaluation 

DXOMARK remains committed to providing reliable professional imaging evaluations for the industry and consumers through scientific, objective, transparent, and repeatable testing methodologies. We will continue to drive innovation in the imaging industry through science. Based on an end-to-end imaging framework, DXOMARK will continue conducting large-scale user preference studies across different regions worldwide. 

We are continuing this collaborative approach and developing a demonstration solution for the full imaging pipeline. The solution incorporates dual-channel, RGB, and spectral sensors as inputs for image tuning and display color calibration. The goal is to establish a standardized framework and extend it to currently under-evaluated stages of the imaging pipeline, ensuring consistent color reproduction and image quality across devices. 

— ams OSRAM, Head of Product Line Ambient Light Sensors, Marcel Knecht 

 At MWC 2026, ams OSRAM showcased a demonstration of its latest sensing technologies and their impact on image rendering at capture level. Using a DXOMARK-validated evaluation scene, the demo highlighted how advanced solutions such as multispectral sensors can significantly improve white balance reproduction by capturing richer and more precise information directly from the scene. 

For more information about the demonstration, please refer to the following video presented by Julia Stuebinger, Head of Marketing – Integrated Optical Solutions at ams OSRAM.

To watch the video with Chinese subtitles, click here