A monitor is not built in one place. The panel comes out of a fab, the module is sealed at one plant, the driver board is populated somewhere else, and the thing you unbox is the end of a chain of assembly steps that each carry their own failure modes. When a brand says “made in our own factory”, the useful question is not whether that is true. It is which of those steps actually happen under the same roof, and which ones are simply passed through.
Our line is run by Jiangxi Shengyu Technology, and the published capacity is around 100,000 monitors a month. That number matters for what it forces rather than for what it proves: at that volume you cannot inspect your way to quality, so the process has to be right by construction. What follows is the sequence, stage by stage, and what each stage is actually there to catch.
The panel is the starting point, not the product
An open cell or a finished OLED module arrives with its own tolerances: brightness uniformity, colour shift across the field, a few dead pixels inside an agreed limit, and a specified range for how much the panel may drift as it warms. None of that is under the monitor maker’s control. What the monitor maker controls is everything that happens after the panel is opened: how it is mounted, how it is bonded, how it is driven, how it is cooled, and how much of the unit is actually tested before it is boxed.
That distinction explains why two monitors with the same panel can feel like different products. It also explains why “we use LG panels” is not a quality statement on its own. A panel is a component. The build is the product.
Lamination: the step that decides how the screen looks at an angle
Bonding a cover glass or a touch layer to the panel is where an air gap either exists or does not. An air gap between the panel and the cover is cheap and it works, but it costs you two things: reflections at the extra interface, and a visibly softer, hazier image at an angle. Fully bonded assemblies remove that interface, which is why they look cleaner in a bright room and why they cost more to build.
There are two common ways to do it. A sheet adhesive is fast and predictable, and it is less forgiving of dust and of any curvature in the cover. A liquid resin cured under UV fills irregular gaps and handles curved surfaces better, but it adds a curing step and an edge-clean step, and the edge clean is exactly where a rushed line leaves residue.
The failure modes are worth naming because they are the ones a buyer actually meets later:
- A trapped particle becomes a bright spot that never moves and never cleans off.
- A bubble at the edge becomes a slow ingress point for moisture.
- An uneven bond shows up as a patch of haze that is only visible on a white field.
- Edge lift does not fail on day one. It fails after a year of temperature cycling.
Polishing and edge work: where cost becomes visible
Cut glass has a raw edge. Grinding and chamfering it is a small step that separates a build that feels considered from one that does not, and it is one of the few quality signals you can check with a fingertip in a shop. A sharp or chipped edge is also a mechanical weak point: cracks in cover glass almost always start at a corner or an edge, not in the middle of the panel.
Polishing is where a line either spends time or does not. It is invisible in a spec sheet and obvious in the hand.
Aging and burn-in: the stage that costs the most calendar time
Finished units are run before they are packed. The reason is statistical: a large share of early electronic failures cluster in the first hours of operation, so a burn-in run converts a failure that would have happened in a customer’s second week into a failure that happens on the line, where it can be reworked or scrapped.
For OLED the same step does a second job. Running the panel exposes uniformity problems that a five-second power-on test cannot see, because some drift only appears once the pixels have been driven for a while. It does not make the panel immune to long-term wear — that is a different, much slower process, which we covered separately in the OLED burn-in explainer. An aging step catches the unit that was going to be bad out of the box. It cannot catch the unit that will be worn unevenly in three years because someone left a bright toolbar on it.
This is also the stage that costs the most floor space per unit of output, because units occupy the line while doing nothing except being on. A plant optimising purely for throughput has an obvious incentive to shorten it.
Calibration: “factory calibrated” is not one thing
A calibration claim can mean at least four different things, and the differences matter more than the wording:
| What the claim can mean | What it actually gives you | |
|---|---|---|
| One profile per model, written once | Consistency of intent, no consistency unit to unit |
| A one-point check at a single brightness | Grey-scale accuracy near that point, and nowhere else | |
|---|---|---|
| A multi-point calibration per unit | Even grey-scale and white point across the range |
| A per-unit calibration stored in the monitor | The result survives a change of computer |
|---|
| Stage | The failure it exists to catch | What you notice later if it is skipped |
|---|---|---|
| Lamination | Dust, bubbles, edge lift, uneven bonding | A bright spot that never moves, or haze at the edge after a year |
| Edge polish | Sharp or chipped glass, coating defects | A crack that starts at a corner |
|---|---|---|
| Aging run | Early driver and panel failures, uniformity drift | A unit that dies in month two |
| Calibration | White point and gamma spread between units | Two identical monitors that do not match |
|---|---|---|
| Thermal soak | Marginal parts that only fail when warm | Flicker or a dropout after an hour of use |
| Port and cable cycle | Connector solder and strain relief | A USB-C port that stops charging after a year |
|---|---|---|
| Drop, vibration, packaging | Internal mounting and carton design | A cracked bezel from a delivery that looked fine |