What Lives Between the Glass
Fourteen distinct layers stand between the LED and your eye. Knowing which one fails, and why, turns a troubleshooting guess into a diagnosis.
The Stack, Bottom to Top
Start with the light source itself. Most modern LCD backlights use a blue-pump white LED — a blue InGaN die with a yellow phosphor cap — arrayed either directly behind the panel (direct-lit) or along one edge with a light-guide plate (LGP) to redirect that light toward the viewer. Edge-lit designs are thinner, but the LGP introduces a constraint: luminance uniformity across the sheet depends entirely on the precision of the microstructure etched or printed into the LGP's bottom surface. Those structures scatter light upward in a calculated gradient, denser with distance from the LED strip, correcting for the natural fall-off as you move away from it. When a display develops a bright bar along one edge — especially after thermal cycling — the LGP's coupling interface has degraded or the LED array has partially failed. It is rarely the panel itself.
Above the LGP sits a diffuser sheet, sometimes two, sometimes one diffuser laminated with a prism film on the same substrate. The diffuser's job is simple: homogenise. Point sources become area sources. The tradeoff is optical efficiency — each diffusion event scatters photons in directions you did not want, and some fraction escapes the system entirely. Prism films (colloquially called brightness-enhancement films, or BEF) partially recapture this by total-internal-reflection redirecting off-axis light back toward the viewer axis, recovering a significant fraction of otherwise lost lumens. The specific angular cone the BEF targets is chosen at design time; panels destined for desktop use optimise for a different cone than those going into a commercial signage installation seen from oblique angles.
Now you reach the first polariser — the rear, or input, polariser. Here the physics becomes unforgiving. A linear polariser transmits, at best, roughly 50 % of unpolarised incident light, absorbing the other half as heat. This is not an engineering failure; it is a consequence of the photoselective absorption mechanism that creates linear polarisation in the first place. Iodine-doped polyvinyl alcohol (PVA) is the dominant polariser material — stretched uniaxially so the iodine chains align, then laminated between triacetate cellulose (TAC) protective layers. The orientation of this rear polariser defines the reference state for the entire electro-optic system above it.
The TFT array comes next. This is a thin-film transistor backplane fabricated on glass, carrying one switching transistor per pixel (or per sub-pixel in some architectures), plus the storage capacitor that holds the pixel voltage between refresh events. The backplane material — amorphous silicon, LTPS, or oxide depending on the panel's refresh-rate and resolution targets — sets the ceiling on electron mobility and, therefore, on how quickly a pixel voltage can be written. Between the TFT layer and the liquid-crystal layer is an alignment layer, typically a rubbed polyimide film or a photo-aligned variant, nanometres thick, whose surface geometry anchors the liquid-crystal director at a defined pretilt angle. Lose that alignment — through contamination or UV damage — and you get persistent domain defects visible as patches of anomalous grey level. They do not appear on a single-colour screen; they appear on gradients, which is why this failure mode is frequently missed until calibration.
The liquid-crystal layer itself is typically a few micrometres thick. In an IPS (in-plane switching) cell, both electrodes sit on the bottom substrate, and the field drives the LC director to rotate within the plane of the glass, minimising off-axis colour shift. In VA (vertical alignment) cells, the director tilts perpendicular to the substrate at rest, producing a near-extinction state that is the source of VA's superior native contrast — typically an order of magnitude better than IPS. The physics of that extinction is worth understanding: when the LC is fully vertical and the rear polariser is crossed with the front polariser, almost no light passes. "Almost" is doing real work in that sentence; small pretilt deviations and imperfect polariser extinction ratios are what set the floor. Every contrast ratio figure you see on a spec sheet is bounded by these values, measured in a dark room with a full black field and a single white window — conditions the panel will rarely see in use.
Above the LC layer is the colour filter array, on the top glass substrate. Red, green, and blue dye-matrix filters — patterned by photolithography, one colour at a time — sit over each sub-pixel aperture. A black matrix of chromium or organic resin blocks the gaps between sub-pixels, preventing light from the TFT metal lines and storage capacitors from leaking through as spatial crosstalk. The colour filter's spectral transmission characteristics are the primary determinant of the panel's native colour gamut; pushing toward wide gamut means narrowing the filter passbands, which costs transmission efficiency, which costs peak luminance. The tradeoff is not free, and any panel claiming both extreme brightness and wide gamut is either using a quantum-dot backlight to shift the source spectrum closer to the filter peaks, or is lying about one of the two numbers.
The front polariser, crossed relative to the rear polariser, completes the electro-optic system. In the off-state LC orientation, it blocks; in the driven state, it transmits. Its failure mode is delamination or yellowing — visible as a warm colour cast at panel edges where the TAC layer absorbs atmospheric moisture and the iodine-PVA complex degrades. This is accelerated by high heat and is a known failure mode in panels operated at sustained high brightness in poorly ventilated enclosures.
Finally, cover glass — or, on mobile panels, a rigid cover lens bonded directly to the front polariser via optically clear adhesive (OCA). That bond layer matters. Air gaps between the polariser and the cover create Fresnel reflections at two surfaces rather than one, reducing contrast in ambient light conditions. Optical bonding eliminates those interfaces at the cost of repairability; a cracked cover glass, once bonded, takes the polariser with it. On large-format panels the cover is often absent, with only a hard-coat on the polariser surface providing scratch resistance — which is why the screen surface of a professional monitor feels different under the finger from a tablet.
What Fails, and Where
Contrast collapse in the dark room traces almost always to the LC or polariser. Uniformity failure traces to the backlight — LGP coupling, LED binning mismatch, or thermal expansion displacing a prism film. Colour accuracy drift traces to the colour filter aging (rare in practice), the backlight spectrum shifting as phosphor degrades, or the front polariser yellowing. Gradient artefacts and domain defects trace to the alignment layer. Knowing the stack means knowing which experiment to run: a dark-room uniformity measurement, a spectral white point measurement, or a field-off dark-level measurement with the panel temperature stabilised.
Fourteen layers. One image. Any one of them can be the culprit.