Backlight Strobing Eliminates Retinal Blur but Halves Peak Luminance

Competitive players chasing the absolute ceiling of mechanical performance will tell you that a 360Hz monitor is functionally useless if the backlight stays on continuously. Display purists, meanwhile, argue that plunging a premium panel into artificial darkness and triggering double-image artifacts just to read a spinning crosshair is a terrible trade.

The debate centers on backlight strobing, a display technology marketed under acronyms like ULMB, DyAc, and ELMB. By violently flashing the monitor's light source on and off, the technique eliminates the motion blur created by human eyes tracking moving objects. But that mechanical clarity comes at a severe physical cost to the panel's peak luminance.[2]

Understanding why manufacturers intentionally turn off the lights on their fastest displays requires looking at how modern screens draw motion. Nearly all contemporary LCD and OLED monitors rely on a sample-and-hold delivery mechanism. The display renders a frame, pushes it to the pixels, and holds that static image perfectly still until the next frame is ready.

That static hold is the root of the problem. When a player's eyes smoothly track a target sprinting across the screen, the physical display is not actually moving the image smoothly. The eye is in continuous motion, but the on-screen object is jumping from one static position to the next, remaining frozen in between.

The Mechanics of Retinal Persistence

Because the human eye is moving while the on-screen image is temporarily stationary, the frozen frame smears across the retina. This phenomenon, known as retinal persistence, creates perceived motion blur even if the monitor's pixels transition colors instantly. A panel can boast a 0.03-millisecond response time, but if the frame is held, the eye will still blur it.

Higher refresh rates attempt to brute-force this problem by simply holding the image for less time. At 60Hz, a frame sits on the screen for a grueling 16.7 milliseconds. Pushing the panel to 144Hz cuts that persistence window down to 6.94 milliseconds, while a bleeding-edge 360Hz display holds the image for just 2.8 milliseconds.

Yet even at 360Hz, that 2.8-millisecond hold time is a non-zero value. The blur shrinks roughly in proportion to the refresh rate, but it never entirely vanishes under a sample-and-hold paradigm. To completely eliminate retinal smearing and achieve the crystal-clear motion of a 1990s cathode-ray tube (CRT) monitor, the display must stop holding the image.

This is where backlight strobing enters the equation. Instead of keeping the pixels illuminated for the entire 6.94-millisecond duration of a 144Hz frame, the monitor's firmware intentionally kills the backlight for the majority of the cycle. The screen goes entirely black, hiding the static frame from the viewer's tracking eyes.

The Luminance Penalty

By inserting a black interval between every frame, the monitor forces the eye to perceive only brief, isolated snapshots of the action. "Without BFI, a 120Hz monitor holds each frame on screen for 8.33ms continuously — creating 8.33ms of pixel persistence," explains display analysis site ScreenHz. Strobing cuts that persistence down to the pulse width.

If the backlight only pulses on for 1.73 milliseconds before going dark again, the effective Moving Picture Response Time (MPRT) drops to 1.73 milliseconds. The retinal smearing is instantly shattered, leaving razor-sharp edges on fast-moving targets. However, that mechanical victory introduces an unavoidable mathematical penalty to the screen's overall brightness.

If a monitor disables its backlight for 75 percent of the frame cycle to achieve that clarity, it is physically emitting light for only a quarter of the time. A panel rated for a peak luminance of 400 nits will suddenly output an effective 100 nits. This severe brightness drop forces a direct choice between motion clarity and usable visibility.[1]

"To achieve backlight strobing, ULMB disables the backlight 75% of the time," noted NVIDIA in a technical breakdown of their original 2015 technology. "This 25% duty cycle on a max 300 nit panel meant that the images would be clear, but less bright." Many players simply turned the feature off rather than squint into the dark.[1]

Modern implementations attempt to compensate by aggressively overdriving the backlight voltage during the brief window it remains on. BenQ's Dynamic Accuracy (DyAc) and NVIDIA's newer ULMB 2 push the LED hardware to its absolute thermal limits during the strobe pulse. This brute-force voltage spike recovers some of the lost luminance, but it cannot entirely bridge the gap.[1][2]

Stroboscopic Crosstalk

Beyond the brightness penalty, strobing introduces a secondary visual artifact known as stroboscopic crosstalk. For a strobe to work perfectly, every pixel on the panel must finish changing to its new color before the backlight flashes on. If the pixels are still in the middle of transitioning when the light hits them, the illusion breaks.[2]

When the backlight illuminates a mid-transition pixel, the viewer sees a faint, duplicated outline of the moving object trailing slightly behind or ahead of the actual target. This double-image artifact often looks worse to competitive players than the native sample-and-hold blur it was designed to replace, as it creates visual clutter around precise hitboxes.

Crosstalk is notoriously difficult to eliminate because LCD pixels do not transition at uniform speeds. A pixel shifting from dark gray to light gray might finish its transition in one millisecond, while a pixel shifting from pure black to white might take four milliseconds. Tuning the strobe timing to perfectly hide all of these varying transitions is an engineering nightmare.

OLED panels, which boast near-instantaneous pixel response times, theoretically bypass the crosstalk issue entirely. Because an OLED pixel can change state in 0.03 milliseconds, the transition is always finished before the strobe fires. However, OLEDs lack a traditional backlight, meaning the panel must turn off the actual emissive pixels to achieve Black Frame Insertion (BFI).

The OLED Brightness Ceiling

Applying BFI to an OLED panel introduces a different set of thermal constraints. Because OLEDs are already limited in their full-screen brightness to prevent burn-in and heat degradation, they lack the luminance headroom to compensate for a 50 percent duty cycle. Turning off the pixels for half the frame plunges an already dim display into near-unusable territory.

Further complicating the setup, backlight strobing is traditionally incompatible with Variable Refresh Rate (VRR) technologies like G-Sync and FreeSync. VRR dynamically alters the frame interval to match the graphics card's output, which wreaks havoc on the fixed, rhythmic timing required for a clean strobe pulse. If the strobe timing drifts, the screen flickers violently.

Consequently, players must usually choose between the tear-free smoothness of VRR and the blur-free clarity of strobing. Only a handful of cutting-edge monitors, utilizing advanced modules like ASUS's ELMB-Sync, attempt to run both simultaneously. Even then, fluctuating frame rates can cause the strobe brightness to waver, introducing micro-flicker that induces eye strain.

By late 2026, hardware manufacturers are split on the best path forward for motion clarity. Some brands are abandoning strobing entirely, betting that pushing raw sample-and-hold refresh rates to 480Hz and beyond will shrink the persistence window enough to satisfy professionals. Others are refining their strobe algorithms, trying to squeeze more voltage into shorter pulses.[3]

For the end user, the decision remains a strict compromise. Enabling a strobe feature guarantees the sharpest possible tracking for fast-moving targets, but demands a dark room and a tolerance for potential crosstalk. Until panel technology can deliver 1,000-nit baseline brightness alongside instant pixel transitions, perfect motion clarity will continue to require turning off the lights.[3]