iPhone 18 Pro Under-Display Camera Pixel Patents

On the iPhone 18 Pro, the screen above Face ID leaves every other row of pixels empty. Under-display camera pixels went from a 1st generation that removes pixels to a 2nd generation that moves the circuits to the edge and lays transparent wiring, and the first to claim that 2nd generation was China's Tianma. But transparent wiring added masks, and Apple chose a 1.5 generation that goes back to removing pixels. Apple itself first described that arrangement in a 2018 specification and drew it in a 2020 filing.

1. Overview

1.1 Above Face ID, every other row of pixels is empty

On the iPhone 18 Pro, released in September 2026, the infrared camera for Face ID moved under the screen. It is an under-display camera (UDC). It sits at the top left of the screen, next to the clock. To the naked eye, it shows up as a barely visible, slightly darker circle.

iFixit looked at that spot under a microscope.

Figure 1. The Face ID camera boundary on the iPhone 18 Pro screen, at 40x. Left is above the camera, right is the normal area. Photo: iFixit

In the normal area on the right, pixels fill the space without gaps. In the area above the camera on the left, rows with pixels alternate with empty rows. iFixit wrote that “over the infrared camera, every other row of display pixels is missing,” and added that below the OLED there is a layer that appears to be “an infrared-pass filter.” 9to5Mac reported that Apple confirmed “the overall pixel density over the module is reduced by half.” To keep the area from looking dim, those pixels are driven brighter.

Remove pixels to open a light path. Simple as that answer looks, the industry went down a different road once before coming back to it. This piece traces that road back through the patent texts. The conclusion comes first.

Under-display camera pixels developed from a 1st generation that removes pixels to a 2nd generation that moves the pixel circuits out to the edge and leaves only transparent electrodes. Surprisingly, the first to claim the 2nd generation was China’s Tianma. But transparent wiring came at the price of resistance and masks, and a Samsung Display filing itself wrote that the wiring could need 4 more masks. Apple chose a 1.5 generation that goes back to removing pixels. Apple first described that arrangement in a 2018 patent specification, and in a 2020 filing drew it as a figure with every other row of pixels left empty. Google, claiming a similar arrangement for cameras in 2021, wrote in its specification a way to do it without changing the deposition mask and without adding a layer.

1.2 Scope of analysis

1.3 Terminology

2. The picture Apple drew in 2018

The idea of removal that the iPhone 18 Pro adopted is already in the specification of a patent Apple filed in April 2018. But it is in the specification, not the claims.

2.1 A grid with every other pixel left out

Apple’s April 10, 2018 filing (US11073712B2) covers an electronic device display for imaging through the screen. Its FIG. 4A looks like this.

Figure 2. Apple US11073712B2 FIG. 4A (priority 2018-04-10). In the central region, every other pixel is left out, staggered from row to row

The specification says this.

There is also a way to halve the density in one direction only. The specification gives as an example a region with “a pixel density of approximately 450 pixels per inch along a length of the display and of approximately 225 pixels per inch along a width of the display,” and explains that adjacent regions like this are alternated “so as to obscure the presence or appearance of lower pixel density.” Halving the density in one direction gives the same shape as leaving every other row empty. FIG. 5A of the specification also shows an arrangement that mixes in groups missing one color (red and green, green and blue) among three-color groups of red, green and blue, operating them together to produce color.

2.2 What it claimed was broader ground

Yet what Apple tied up in claim 1 is not this arrangement.

Lower the pixel density only above the camera and send light between the pixels. In this piece’s division, that is the 1st generation. The use case also centers on light reflected from a touch input, that is, under-display fingerprint imaging. In its claims, Apple’s 2018 filing stands in the 1st generation; in its specification, it had already drawn the 1.5 generation. In between, the industry made a trip down a different road called the 2nd generation.

3. The three branches at a glance

Figure 3. The three branches compared on the same screen and the same camera position (dashed circle). The 1st generation spaces pixels out and sometimes enlarges them for lifetime. The 2nd generation moves circuits outside the circle and connects them with transparent wiring. The 1.5 generation leaves every other row of pixels empty and keeps the circuits in place. Sizes and spacing are not to scale

The reason for calling it the 1.5 generation is in Figure 3. On transmittance alone, the 2nd generation has the clearest advantage. Yet today’s product has come back to a point closer to the 1st generation than to the 2nd.

The model below lets you switch between the three branches.

4. 1st generation: removal

The earliest form of the idea of sending light between pixels, within the range this piece confirmed, is a 2009 Sony filing.

4.1 Sony 2009, a window at every pixel

Claim 1 of the filing Sony made with a November 2, 2009 priority date (US20110102308A1) is short.

Open a small window at each pixel position and image from behind it. Google Patents records show this one lapsed for nonpayment of maintenance fees.

4.2 Sparse only above the camera

In 2018, the idea narrowed to the area above the camera. Apple’s claim 1 (2018-04) is one such case, and BOE’s June 2018 filing family (US12266309B2) also lowered pixel density only in the area above the camera. BOE’s August 2018 filing (US11037994B2) placed transmissive parts between pixels. These two were covered in CLAIM’s BOE and Samsung Display under-panel camera patent suit analysis.

Google’s May 2019 filing (WO2020243740A1) added two things. It rearranged the circuits in the reduced-density area, and laid a light-blocking layer to prevent diffraction from the gaps between circuits. If the gaps are regular, light bends at them and produces blur and light streaks in photos. TEKER’s Why does a selfie shot through the screen come out fine? covered this problem in detail.

4.3 Enlarge pixels by as much as you remove: lifetime

Removing pixels makes the area above the camera darker. To produce the same brightness, more current has to flow through the remaining pixels, and that wears them out faster. The specification of Tianma (Wuhan)‘s October 2019 filing (CN110752242B) wrote out this chain exactly.

So the pixels in the removal area are enlarged. A larger emitting area can produce the same brightness at a lower current density. The specification of LG Display’s 2019 filing also made the reduced-density pixels above the camera larger than those in the normal area: “a size of each of the transparent area organic light emitting diodes TOLED may be set to be greater than that of each of the opaque area organic light emitting diodes OOLED provided in the opaque area.”

This size comes up again in Section 7.

4.4 The limit of the 1st generation

Even after removal, the remaining pixels still have circuits attached. The specification of LG Display’s 2019 filing stated that limit briefly.

Removing pixels alone leaves the circuits behind. Then why not take the circuits out from above the camera altogether? The 2nd generation came out of that question.

5. 2nd generation: circuits to the edge, and China came first

Place the pixel circuits outside the camera area, and leave only transparent electrodes and the emitting layer above the camera. Within the range this piece confirmed, the first to claim this concept was China’s Tianma (Wuhan), on June 28, 2019. That is half a year earlier than LG Display’s filing.

5.1 The prototype was the transparent display

The thinking behind the 2nd generation came from transparent displays. Tianma’s October 2017 filing (CN107768407B) divided the sub-pixel of a transparent display into a transmissive area, an opaque area, and an emitting area, and placed the driving circuit only in the opaque area. Gathering the circuits to one side is what makes the rest transparent. It is not for cameras, but it is the prototype of the 2nd generation.

5.2 Tianma 2019, only transparent electrodes above the camera

One year and eight months later, Tianma brought the same idea to the area above the camera. It is CN110288945B, with inventors Peng Chao (彭超) and Cui Ruili (崔锐利).

The specification also gave the reason. Placing the pixel circuit in the non-display area reduces the light-blocking area above the camera and raises transmittance. It named ITO, IZO and IGZO as transparent electrode materials. The display unit density was kept the same as in the normal area, so only the circuits were removed without sacrificing image quality. This was filed only in China.

5.3 LG Display 2019, transparent area electrode lines made of ITO alone

Half a year later, on December 24, 2019, LG Display filed in the same direction (US11751463B2, inventors JeongHyeon Choi, Sul Lee and Eunil Cho). It moves the circuits that drive the pixels in the transparent area above the camera to an outer area and connects them with transparent area electrode lines. The specification wrote that “the transparent area electrode line TEL may include only the transparent metal,” and made the cathode of transparent metal only as well. It is this transparent area electrode line that went into the representative claim.

5.4 Samsung Display was looking elsewhere at the same time

There is a Samsung Display filing from June 17, 2019 (KR102832798B1), 11 days earlier than Tianma’s. It lowered the density above the camera and placed around it a transition region where density rises gradually with distance. It is a 1st-generation refinement to keep the edge of the removal area from standing out. It has no concept of moving circuits or connecting with transparent electrodes. That is why this piece takes Tianma’s June 2019 filing as the first 2nd-generation claim.

6. The price of the 2nd generation: resistance and masks

Transparent wiring saved the light, but it paid two prices: resistance higher than metal, and additional masks to pattern that wiring. Both were written into the specifications by the very companies that claimed the 2nd generation.

Figure 4. Process burdens written in the patent texts themselves. None of the documents states a cost figure

6.1 Resistance: the farther, the weaker the signal

Tianma had already pointed out the problem in its original filing. With the circuits at the edge, connection lines have to be run into the pixels deep inside the camera area.

LG Display matched that resistance with line width. The longer the electrode line, the higher the resistance, so the line width is set differently according to length to reduce differences in resistance and voltage drop between lines. Table 1 in the specification shows that when every line width is the same 5 micrometers, resistance spreads from 0 to 10 megaohms across sections depending on line length. It means each transparent line has to be designed separately for its length.

6.2 Masks: 4 more

The process-side price was counted by a Samsung Display filing. The February 1, 2021 filing (KR20220111767A, inventor Woo Min-woo et al.) is a 2nd-generation structure. It places the pixel circuits for the transparent area in an intermediate area next to the transparent area, and connects them to the anodes in the transparent area through three layers of transparent connection lines (TCL1, TCL2, TCL3) made of ITO, IZO and IGZO. It then wrote out the process for forming each connection line, one by one.

FIG. 14 of this filing lists the normal-area process with mask numbers on the left, and draws the additional transparent connection line processes separately on the right. Each extra mask adds another round of exposure, development and etching. For a camera in one corner of the screen, four rounds are added to the process for the entire panel.

7. 1.5 generation: removal again, with the same layers

While the 2nd generation kept adding transparent wiring, filings appeared on the 1st-generation side that refined the removal method. In 2020 Apple drew a figure with every other row of pixels left empty, and in 2021 Google claimed an arrangement that neither changes the deposition mask nor adds a layer. Neither moves the circuits nor lays transparent wiring.

7.1 Apple 2020, every other row left empty

Two years after the 2018 filing in Section 2, Apple filed “Displays having transparent openings” (EP4193388B1 · CN115956357B, inventor Yuchi Che et al.) with priority to a US provisional application of August 10, 2020. The drawing closest to the product in Figure 1 is in this document.

Figure 5. Apple EP4193388B1 FIG. 6 (priority 2020-08-10). Red, green and blue are removed color by color (612, marked X) from the normal arrangement (610), creating horizontal rows with no pixels (615) and rows of pixels in what the specification calls a zig-zag arrangement (614). The drawing is taken from the Chinese grant CN115956357B and rotated 90 degrees for readability

The specification describes this figure as follows: “the subpixels are removed such that there are horizontal stripes of empty pixel regions.” The remaining emissive sub-pixels “have a horizontal zig-zag arrangement following the pixel removal scheme.” It is the same shape as the “every other row of pixels missing” that iFixit saw.

There is also a figure on the circuit side.

Figure 6. Apple EP4193388B1 FIG. 9. Thin-film transistor circuits are gathered into a strip in one row (74), and the next row (72) is empty with no circuits. The emissive pixels (62) are laid across them in a horizontal zig-zag

The specification explains it this way: “every other row (e.g., row 72) includes no thin-film transistor sub-pixels and every other row (e.g., row 74) includes a continuous strip of thin-film transistor sub-pixels.” The circuits are not moved outside the camera area. They are gathered into one row, and the next row is left entirely empty. Instead of laying new transparent wiring as in the 2nd generation, it removes pixels as in the 1st generation, but organizes the removal row by row.

But what this filing tied up in claim 1 is not this arrangement. Claim 1 is a device that places two sensors under transparent openings with different patterns and uses the image from one sensor to remove the diffraction artifacts in the other sensor’s image. The row-by-row arrangement is in the specification and drawings.

7.2 Google 2021, one in four with a single color

One year and four months after Apple’s 2020 filing, Google filed on December 20, 2021 (US12213364B2, inventor David Morris Hoffman et al.). It is a sub-pixel arrangement for an under-display sensor.

Figure 7. Google US12213364B2 FIG. 1 (priority 2021-12-20). The surrounding area (120) has the normal arrangement, and pixels are removed in the central sensor area (150)

Two pixels in four cells. The density is halved. The specification adds three things.

First, the wiring is reduced to match where pixels were removed. “Data lines and rows that drive the sub-pixels of the third color may be reduced by 50% using staggered connections to emissive pixels.”

Second, the metal data lines crossing the empty positions are not kept straight. The specification wrote that the data lines bending between pixels “follow a non-linear path that reduces flare” (patent FIG. 5). Straight lines in a regular row bend light like a diffraction grating. Bending the lines breaks up that regularity. It solves the diffraction problem from Section 4 with the shape of the wiring rather than a light-blocking layer.

The third is where it parts ways with the 2nd generation.

The 2nd generation opened the window by patterning additional transparent wiring layers. The 1.5 generation opens the window by not patterning part of the metal layers that were already there. The masks the two documents talk about are of different kinds. What Samsung Display counted as “four additional” are photomasks for patterning wiring layers, and what Google says is unchanged is the deposition mask for the emitting layer. But the direction clearly splits. One side adds layers, and the other does not.

Pixel size splits too. As seen in Section 4.3, the removal area had the option of enlarging pixels to protect lifetime. Google instead tied size down in claim 1: the three-color pixels above the camera occupy “a same amount of space within the grid of pixel-sized regions” as the three-color pixels in the normal area. According to the 9to5Mac report, the iPhone 18 Pro took the route where “Apple increases the brightness of these pixels to compensate” rather than enlarging them. How lifetime was handled at that price could not be found in public documents.

7.3 The documents Google stands on

The Google patent cites 9 documents. Those the applicant cited directly include Apple’s 2018 publication (US20190310724A1, that is, US11073712B2 from Section 2) and Google’s own 2019 filing, listed together with BOE’s 2018 filing and 4 Samsung Display documents (3 of them cited by the examiner). Google, in effect, claimed for cameras as a 2×2 arrangement, three and a half years later, the “omitting every other pixel” and the “incomplete subpixel groups” missing one color from the Apple specification seen in Section 2.

Tianma’s 2019 original 2nd-generation filing is not among these 9.

7.4 Apple’s choice

Back to Figure 1. The screen above the camera on the iPhone 18 Pro halved its pixel density, and the remaining rows of pixels still carry all three colors and their circuits. It is not the 2nd generation with circuits moved to the edge, but the 1.5 generation that goes back to removal.

The 2nd generation has the best transmittance. But transparent wiring brings resistance and, as the Samsung Display filing itself wrote, raises the mask count by up to four. Amid that burden, Apple chose the 1.5 generation. And that arrangement points in the same direction as what Apple first described in its 2018 specification and drew as a row-by-row figure in its 2020 filing.

Figure 8. Priority dates of the three branches and the iPhone 18 Pro. Apple's 2018 filing appears twice. Its claim 1 is 1st generation, while its specification is already 1.5 generation

8. Counterevidence and limitations

This piece did not determine which patent the iPhone 18 Pro practices. Whether it is Apple’s own patent, whether it has any relationship to the Google patent, and who made the panel could not be found in public documents. What this piece says goes only as far as this: the product uses a 1.5-generation-type arrangement, and that arrangement was first written down in Apple’s specifications (2018 and 2020) and Google’s claims (2021).

No document proves that “they moved to the 1.5 generation because of masks.” The resistance and mask burden of the 2nd generation, and the statements that the 1.5 generation adds no layer and keeps the deposition mask unchanged, are all in the patent texts. But no company wrote “so we abandoned the 2nd generation.” “Apple chose the 1.5 generation” in Section 7.4 is this piece’s reading, placing those facts side by side.

This piece does not attribute intent to design around. It is a fact that Tianma’s original filing is not among the documents cited by the Google patent. But whether that was because Google did not know of it, avoided it, or was solving a different problem in the first place cannot be told from the documents.

The finding on the 2nd-generation originator holds within the range checked. A Chinese-language filing that does not show up in English searches could be earlier.

The 1st, 2nd and 1.5 generations are this piece’s division. They are not names the patent texts use for themselves, nor an official industry classification.

The drawings that most resemble the product are FIGS. 6 and 9 of Apple’s 2020 filing (Figures 5 and 6 in this piece). The Google claim places, within each group of four cells, a pixel that keeps only one color, so its arrangement differs from the product. Apple’s 2020 filing also wrote that arrangement in the specification, not the claims. This piece does not say the product was made according to these drawings. There is no material to compare it sub-pixel by sub-pixel.

9. Conclusion and implications

The history of under-display camera pixels is not a road that only went forward.

The 1st generation removed pixels. The 2nd generation moved the circuits to the edge and laid transparent wiring to push transmittance as far as it would go, and its first claim was China’s Tianma in June 2019. But transparent wiring came at the price of resistance and masks, and a Samsung Display filing put that price as “four additional masks.” The answer Apple put out in 2026 was to go back to removal. The first picture of that answer was in Apple’s own 2018 specification, and Apple’s 2020 filing went as far as drawing a figure with every other row of pixels left empty. In 2021 Google claimed an arrangement for cameras that adds no layer.

The structure with the best transmittance does not necessarily become the mass-production structure. In the texts read for this piece, what separates the 2nd generation from the 1.5 generation was not the pixel arrangement drawing but the process table.

Two things are worth watching. One is how claims among Apple, Google and the panel makers overlap around the 1.5-generation arrangement. Filings that vary slightly where pixels are removed and how the wiring is shaped keep piling up, and there is also a 2023 Samsung Display filing (US20240407209A1) in which the examiner cited the Google patent. The other is whether the 2nd generation will disappear. If a process comes along that lightens the mask burden, the 2nd generation, ahead on transmittance, could return. When that happens, the side holding the original claim is China.