Why Megapixels Matter: Measuring High-Resolution, Pixel-Dense Displays
For years, the trend in displays has driven relentlessly towards higher-resolution screens that feature an increasing number of smaller pixels, more densely packed together. Digital imagery went mainstream in the 1990s, and LCD screens with LED backlights took off in the early 2000s; sales of LCD screens overtook older cathode ray tube (CRT) televisions in 2007. Since then, the emergence of new technologies such as OLED and microLED have helped accelerate the trend. HD (high definition), 4K, and 8K display devices have brought high-resolution viewing experiences to the mass market.
Projected sales of high-resolution 4K television screens (which, along with 8K screens, are called ultra high definition or UHD) continue to grow around the globe. (Image: IHS Markit)
These high-resolution screens aren’t cheap, however, and consumers expect high performance in return for a high sticker-price. Consumer electronics device manufacturers are tasked with ensuring their display devices deliver a flawless visual experience with bright, crisp, and defect-free screens.
When it comes to display quality, manufacturers have long relied on automated visual inspection solutions (AVI, also called automated optical inspection, AOI) such as cameras and imaging systems. Using these tools, each display panel can be quickly inspected to spot defects (such as dead pixels) and uniformity issues (or mura, from the Japanese word for “blemish”). Among display metrology providers like Radiant, there has been a parallel trend towards higher resolution specifications for inspection systems. Put simply, it takes a high-resolution system to measure a high-resolution display.
High resolution is important not only for big screens like televisions, but also for small displays such as smartphones and smart watches that are viewed close up. Users expect a crystal-clear image. Shown here: the Apple Watch 5 and iPhone 11 Pro.
Are We Talking About Pixels? Or Pixels?
The term “pixels” gets used often, but it can refer to different things. The word pixel is derived from “picture element” and is used to describe the individual illuminated elements of a display screen, or the individual ink dots of a printed image. In photography, a pixel refers to the smallest element of a camera’s image sensor (a photosite) that can record the light (photons) that enter the camera when the shutter opens. The sensor pixels store an image in digital form by converting photons to electrons.
Display inspection applications are concerned with two specific types of pixels: display pixels and image sensor pixels. When the consumer electronics industry talks about displays (such as smartphone, laptop, or television screens), a common specification is the number of display pixels.
For example, a display device listed at 1520 x 720 means the screen is a grid that has 1520 pixels horizontally and 720 pixels vertically. Multiplying 1520 x 760 = 1,094,400 pixels in the entire display. Typically, more pixels = higher resolution = better image clarity and display quality.
Over the years, display makers have been able to shrink the size of display pixels and squeeze them closer together. This is referred to as decreasing the pixel “pitch” (the spacing between pixels), which creates a denser array of pixels and a higher-resolution visual image on screen.
Looking at different display panels in the photograph below, notice how the “dots” (display pixels) in the lower screen (A) are larger, and the black space between them is more pronounced. This display has lower pixel density, fewer pixels per inch (“PPI”), or a larger pixel pitch—these descriptions are roughly equivalent in meaning. The pixels are not very close together, so you can’t fit as many of them into a given area.
Examples of display panels with different pixel sizes, pixel pitches, and pixels per inch (PPI)—all of which impact display resolution.
The other displays in the above image show the pixel size and the space between the pixels shrinking as they progress backwards. The top/back-most panel (B) has the tiniest pixels, spaced the closest together (smallest pixel pitch), and therefore this display provides the highest PPI and the highest resolution.
Imaging systems like Radiant’s ProMetric® cameras are used to measure and inspect displays for quality control during production. Each camera has a certain MP (megapixel) specification—the greater the number of pixels, the higher the resolution of the camera. For instance, with our latest product launch we now have 45MP and 61MP models (with 45 million and 61 million sensor pixels, respectively).
Not to be confused with display pixels, however, when we talk about the MP of a ProMetric imager, we’re referring to the number of sensor pixels on the camera’s image sensor. While display pixels emit light, sensor pixels capture light and translate it into digital information that we use to measure light or color values.
To provide the most accurate evaluation of today’s high-resolution, pixel-dense displays, it’s important to be able to capture high-resolution images for measurement and analysis. To do this, having more sensor pixels in your camera can make a big difference.
A high-resolution camera can dedicate a larger number of sensor pixels to measure finer details in a display. If a camera has an equal or greater number pixels than the display, the camera’s images can be used to distinguish and measure individual display pixels. A low-resolution camera that has fewer pixels than the display will have to image multiple display pixels with each sensor pixel, losing the ability to distinguish individual display pixels in the image and reducing measurement precision.
A high pixel count translates into high spatial resolution, which determines a system’s ability to distinguish fine detail within an image. For example, the image at far left has a low pixel count (low resolution), so there is little discernable detail. The image at far right has a high pixel count (high resolution), providing excellent detail and clarity.
Why Sensor Megapixels Matter
Below is a conceptual illustration of display pixels vs. camera sensor pixels. In this simplified schematic, each square represents a single pixel. If the display panel (A) and the camera sensor (B) have the same number of pixels (a 1:1 ratio), the data from each display pixel can be captured fully by a single pixel on the camera’s sensor.
Conceptual illustration of display pixels and camera sensor pixels.
However, if the display (C) has many more pixels than the camera sensor (B), it means each sensor pixel must store the data for multiple display pixels (a ratio of 4:1), limiting the amount of detail captured. A low-resolution camera could take multiple images of the display to capture more detail, but this slows the inspection speed (increases measurement or cycle time), which doesn’t work for high-speed manufacturing lines. Ideally, a high-resolution display (C) would be imaged by a high-resolution sensor (D), with several sensor pixels dedicated to capturing the data for each display pixel—in this theoretical example, a 1:4 display-to-sensor-pixel ratio.
High-resolution imaging is also extremely important for measuring each display pixel’s constituent parts. In today’s OLED and microLED displays, each pixel is made up of three subpixel elements (red, green, and blue diodes) that each produces its own light. Because light emissions can vary per subpixel (causing visible quality issues in the display), it’s crucial that a display measurement system capture detail down to the subpixel level. In an ideal case, the measurement system should capture this level of detail in a single image (a single camera shot) to keep up with the pace of manufacturing production lines.
Magnified view of OLED display pixels, each containing one red, one blue, and two green subpixel elements.
New High-Resolution Inspection Solutions: ProMetric 45MP and 61MP
At this year’s SID (the Society for Information Display) 2021 Display Week Symposium & Expo, Radiant announced a new generation of high-resolution systems: the ProMetric Y45 and Y61 Imaging Photometers and the ProMetric I61 Imaging Colorimeter. The “45” and “61” in the product names refer to the number of megapixels in each camera’s sensor.
Increasing our camera resolution allows us to apply more sensor pixels per display pixel, increasing the amount of detail that our cameras can capture—for example, allowing us to measure individual subpixels. In addition, Radiant has developed software algorithms and other techniques to maximize the performance of our cameras for distinguishing and measuring display pixels and subpixels with repeatable accuracy.
Radiant’s product development team is continuously challenged to engineer our systems to strike just the right balance of imaging detail and imaging speed to ensure accuracy and efficiency for display inspection. This means producing higher-resolution imaging systems and more advanced image processing methods to keep up with the increasing resolution of emerging displays.
Both our spaced-pixel measurement method (US Patent 9,135,851) and our fractional pixel measurement method (US Patent 10,971,044) were developed to address the industry’s increasing display resolutions using existing hardware (for more information on these methods, read our White Paper). These new 45MP and 61MP imaging solutions enhance a user’s ability to image the highest-resolution displays with even greater detail and accuracy—while capturing an entire display in a single image for measurement at the speed of production.
Radiant’s new ProMetric I61 Imaging Colorimeter (left) and ProMetric Y45 Imaging Photometer (right).
All Pixels Are NOT Created Equal
Not all sensor pixels are created equal, however. Pixel size and capacity (“well depth”) determine how much data (how many electrons) the pixel can hold, limiting the precision of each sensor pixel’s measurement. There are other qualitative differences in sensor pixels such as dynamic range and signal-to-noise ratio (SNR)—a measure of how much good signal (electrons from the display image) versus noise (electrons from other sources) each sensor pixel captures. Poor dynamic range and a lot of image noise limits measurement accuracy.
Radiant carefully tests available sensor components for these and other criteria before choosing sensors for our ProMetric systems. And then we rigorously text each sensor before, during, and after assembly of our cameras to ensure they are performing up to our high standards—and the high standards our customers have come to expect.
Of course, there are many other factors besides the sensor that contribute to the superior quality of Radiant’s imaging systems including:
- Design and engineering of the electronics to control temperature and limit thermal noise.
- Lens selection and optical design to identify and eliminate issues such as stray light.
- Lens calibration to eliminate distortion or “fisheye” effects from wide-angle optics.
- Integration of photopic and tristimulus color filters with advanced calibrations for accuracy to light-measurement standards (CIE color-matching functions).
- The industry’s leading image analysis and automation software, TrueTest™.
One of the highest-resolution displays available today: Samsung offers 33 million pixels across its 65” NEO QLED screen. (Image © Samsung Display)
By combining high-resolution sensors with careful engineering, Radiant’s new ProMetric imaging systems can perform inspection of high-resolution, pixel-dense displays at high speeds to keep pace with manufacturing production lines. With the new 45MP and 61MP ProMetric models in our product portfolio, we are positioned to solve measurement challenges for display manufacturers now and into the future as microLEDs, quantum dots (QD), and other nano-scale display pixel technologies become increasingly common.