Reading a text from a book vs. on a computer screen is visually identical.
Current LCDs have high enough resolution and contrast so that their image is the same as that of print material — as measured by all commonly used metrics.
So why do so many people experience discomfort only from digital displays and not books?
Digital eye strain is experienced by the majority of computer users and it is debilitating for a small minority of the population. (They have no trouble reading a book.)
Worse yet, there are known negative health effects of digital display use — affecting everyone, also those who do not experience discomfort.
It's the light!
If luminance and contrast are adjusted optimally and the display is placed in an ergonomically correct location, the situation is only different from reading a book in the light emitted by the display.
But that is quite a difference. Here's why.
1) Unnatural spectral composition
Our eyes are easy to fool. The white color of an LED (like the backlight in most computer screens) has a very different spectral composition from daylight of the same color temperature. But the receptors of the retina cannot map the whole spectrum: they only sample from a few places and estimate the rest, more or less.
In natural circumstances, this is good enough, since all natural light sources are black-body radiators: their light emission follows the same physical rules. So, at a certain color temperature, their spectral composition will be identical.
Natural light sources produce light with a continuous spectrum, dependent only on the source's temperature. Their physiological effects, including the balance between tissue stress and recovery, are always in optimal balance for the location and time of day.
Not so with LEDs. The regular LCD monitor backlight consists of a blue LED and a phosphor, which converts part of the narrow-band blue light into a broader band of orange. The two together give white light, with an overabundance of some blue wavelengths and a complete lack of deep red and near-infrared.
A regular white LED emits much of its light right in the range where the retina is most susceptible to photochemical harm. It also lacks the frequencies responsible for aiding tissue recovery.
This leads to the following problems:
Sleep disruption when digital displays are viewed in the last 2-3 hours of the day
Excessive tissue stress, which can contribute to an accelerated aging of the retina and various kinds of eye problems, as well as discomfort and fatigue
Horrendous color rendering, when the "blue spike" is filtered
2) Flicker
LEDs are most often dimmed through pulse-width modulation: a method where the light source is turned on and off in a rapid fashion. While at higher frequencies, this is usually imperceptible, on the level of the retina and large parts of the brain, there is a phase-locked neural response to the flickering, leading to fatigue and headache. Eye movement planning is also disrupted, making eye movements less precise and thus decreasing reading speed and further inducing fatigue.
3) Lack of circadian dynamics
Naturally, light intensity and color temperature change together, continuously through the day. These changes give cues to our circadian clocks and physiologically prepare all tissues for the upcoming phase of the day. While software-based attempts to introduce these cues to digital displays can somewhat imitate the color changes (at the expense of color resolution), the physiological effects cannot follow, due to spectral limitations.
Solution?
There is no shortage of products which aim to mitigate the harm from computer displays. Since the role of blue light in both keeping us awake and causing oxidative stress is quite well-known already, various forms of filtering are available to reduce this short-wavelength contribution.
Owing to the inherently choppy spectrum of the display backlight, however, these either achieve little (all software-based filtering and most forms of physical filtering), or they severely distort colors. When physical filtering is done effectively (the short end of the spectrum is mostly cut off), only a broad band of orange light remains — reducing color contrast so dramatically that all colors, even red, will turn into shades of orange... With full-spectrum light, the same filtering results in a far lower degree of color distortion.
This is also where more expensive and "eye-care" monitors can perform somewhat better: if the blue spike in the spectral composition of the backlight source is less prominent and the total spectrum is somewhat wider, the above color distortion will be less obvious. Similarly, some of these monitors flicker at higher frequencies, or not in all settings. However, they are all based on LED backlighting, inherently carrying the same issues, just to a lesser degree.
Alternatively, some products attempt to "correct" the LED spectrum by supplying some of the missing deep red and near-infrared frequencies from an additional lamp, placed next to the computer screen. While these can have beneficial effects for the eyes, their light will not fall on the most precious central parts of the retina: since the viewer's gaze (foveal vision) remains on the display, it is still primarily affected by its light. These lamps also cannot rid the flicker and the sleep-disrupting effects of blue light, or add circadian dynamics to the light environment.
Finally, e-ink displays (used in daylight) are great for the eyes — when they are able to serve as an alternative. Their lack of colors and inability to display dynamic content, however, make them insufficient for most tasks, like browsing the web, video calls, watching a movie, and so on.
Instead of mitigating harm, Pixun solves the root cause of the problem by using natural light only.
Displays are the only lighting application where the intended use is staring directly at the light source. Therefore, it is paramount that we use the best light available. This is always natural light: a black-body radiator emitting full-spectrum light, tuned optimally for the time of day and location, and without any flicker.