Colorimetric standards do not ensure identical perception

Two televisions can comply with the same colorimetric standards without their images appearing identical to all viewers. A scientific study published in the journal Optics Express at the beginning of October explores how certain modern display technologies can accentuate these differences in perception.

Color accuracy in television testing

Color accuracy is a key focus in television testing protocols, with measurements such as color temperature, Delta E, and coverage of colorimetric spaces used to verify whether a screen meets industry standards. However, even with a perfectly calibrated white point on D65, two viewers may not necessarily see the same thing.

Observer metameric and modern display technologies

This phenomenon was studied by Andrew Rider and Andrew Stockman from University College London (UCL), along with John Frith, in a publication titled Why the colours of modern displays can look so different to different observers. Their work focuses on observer metameric, which takes on particular importance with certain modern display technologies.

Variations in human sensitivity to light

To reproduce colors, a screen combines red, green, and blue light emissions. However, not all individuals are equally sensitive to these. Variations in photoreceptor cells in the retina, related notably to genetics, and the aging of the crystalline lens, which changes the amount of blue light reaching the retina, play a role.

Modern technologies and visible perception differences

These differences existed with older display technologies, but cathode ray tube televisions used phosphors with emissions generally less narrow than those of RGB lasers or certain quantum dot sources. Variations in sensitivity between observers tended to produce less pronounced perception differences. With modern technologies, the phenomenon can become much more visible.

Research methods and simulation results

To study these differences, the researchers used a physiological model of human vision applied to 16 display devices, including televisions, professional monitors, and cinema projectors. Their simulations show that certain variations in sensitivity to red and green, as well as the aging of the crystalline lens, can lead to significant perceptual differences.

Limitations of the study

However, a reservation remains: the predictions of this model have not yet been experimentally validated among real viewers under the studied conditions.

Perfect calibration and perception differences

A perfect calibration does not guarantee identical perception. Should we conclude that our colorimetric measurements are no longer useful? Certainly not. They remain essential for evaluating a television's color fidelity and verifying that it meets the intentions of content creators.

Standardized references and individual differences

However, these measurements rely on standardized colorimetric references, built from standard observers. The CIE 1931 observer remains the historical and largely dominant reference for colorimetric measurements in the video industry. It does, however, not perfectly represent the visual sensitivity of every individual.

Perceptual differences in real-world examples

Imagine two televisions using different technologies, both with a white point measured as identical to D65. Some viewers might perceive a slight pinkish tint on one and a greenish tint on the other. The instruments did not necessarily make an error: the two screens produce different light spectra, to which our eyes do not all react in the same way.

Calibration and personalized corrections

This is an important distinction for interpreting the results of our tests. A very low Delta E indicates that a screen reproduces colors accurately according to the reference used. It does, however, not guarantee identical perception for all viewers. Could we then calibrate each television according to the eyes of its owner? In theory, personalized corrections could reduce certain differences.

Innovations in display technology

The researchers also mention the possibility of multiplying primary light sources to better control their emission spectrum. This approach echoes recent innovations by Hisense, with its four-component RGB-C, including a cyan component, or its MicroLED RGB-Y incorporating yellow. These technologies, however, have not yet demonstrated their ability to reduce observer metameric.

Short-term recommendations and industry standards

In the short term, however, there is no justification for arbitrarily changing the settings of one's television. Colorimetric standards remain an essential common reference for accurately reproducing films and series.

Perception and image quality

This diversity of perception is not unknown to home theater enthusiasts. Some viewers are particularly sensitive to speckle, this buzzing caused by light interference from laser projectors, or to the rainbow effect in certain DLP models. These phenomena are based on different mechanisms from metameric, but illustrate the same reality: the quality of an image is not solely determined by what an instrument can measure, it also depends on how each person perceives it.

Limits of calibration

This study reminds us of a fundamental limit of calibration: we can measure with great precision the light produced by a screen, but this does not guarantee that all eyes will interpret it in the same way.