Performance Optimization for In-Vehicle Displays Performance Optimization for In-Vehicle Displays

Performance Optimization for In-Vehicle Displays

As the number and size of displays in cars continue to increase, harmonizing those displays across different applications and technologies within interior surfaces is a key task for system integrators. High-performance displays in vehicles require adjustments for specific design and performance needs. This article covers aspects of premium system design that involve optimization of white-point adjustment, color, and black uniformity for single and multiple display applications.

by Kai Hohmann and Markus Weber

Displays play a major role in automotive interior design and are destined to play an even larger role in the future, as advanced driver assistance systems (ADAS) and automated driving technology will require displays in ever greater numbers. The displays used in these systems will be installed in close proximity to each other, which necessitates the harmonization of their visual quality and appearance. Yet displays vary in terms of position, technology, display vendor, size, and purpose. Specific issues that this article addresses are white-point adjustment, black uniformity, and color adjustment – all of which gain in importance as they influence the perception of the vehicle’s interior quality.

System Integration Tasks

Automotive displays provide the biggest share of the visuals within a holistic interaction environment. Typically, due to sourcing strategies, these displays come from various manufacturers, are based on different technologies, and are used for different applications such as cluster instrument, center stack, and rear-seat entertainment. Increasingly, displays are placed in prominent positions without light-shielding hoods, so that the impact of the ambient light dynamics typical in automotive applications increases. High contrasts of >1,000:1 make any deviation in black uniformity, especially at night, clearly visible. In-vehicle displays also often include touch capabilities and are more and more often curved, with a bonded 3D surface. Front-cover material options include glass and plastic, plus coatings. Designs that increasingly include seamless integration and free forms, as shown in Fig. 1, further add to the integration challenges.

It should also be kept in mind that it is not only the displays that need to be harmonized in appearance among themselves; they also need to blend in with other illuminated elements of interior design, such as switches and buttons. Obviously, the most sensitive challenge is to ensure a uniform night design.

In order for these multiple displays to perform adequately under one surface, such as a dashboard, as shown in Fig. 1, each needs to be calibrated very carefully and optimized on a system level – one that is inclusive of all device components. To achieve a uniform high-quality appearance and seamless integration into an interior concept, the authors’ company, Continental, uses mature algorithm-based processes for 100 percent end-of-line measurement and adjustment of key parameters such as white point and primary colors. Note: “End of the line” in this case means at the Tier 1 supplier (Continental) before the display unit is shipped to the OEM (the car manufacturer). This focus reflects a strong position in the automotive display market: According to Q1 2017 IHS automotive market data, Continental is the number one buyer of displays for cluster instruments and among the top three buyers of displays for center stack applications.a The sheer number of displays and the scope of use cases make mature measurement and adjustment procedures a core integration task for system integrator companies like Continental.b

Fig. 1:  This prototype automotive dashboard with six OLED displays provides a good example of the type of seamless display integration and new form factors that car manufacturers are seeking to implement. Source: Continental Automotive GmbH.

White-Point Adjustment

A commonly accepted method of white-point adjustment is realized by recalculating the original image data at runtime by applying different scales for red, green, and blue luminance. The maximum scales of red, green, and blue image data are fixed during production using a complex calibration procedure. The following practical look at white-point adjustment from the perspective of a large-scale display buyer for automotive applications shows the scatter range of display parameters and thus the relevance of the adjustment processes. The statistics provided are based on one type of many in-plane switching-based LCDs of different sizes and reflect a population of >100,000 measured units. The white point of each display in this group was measured from two driver viewing perspectives: for the instrument cluster display, this was a perpendicular view, while for the center stack, the view was from a 30° sideways angle. Color coordinates in x/y and luminances of white, red, green, blue, and black were measured. From these parameters the target white point and target primary colors were calculated in an iterative algorithm. Figure 2 shows the established adjustment principle in the typical CIE graph.

Fig. 2:  The CIE 1931 color space graph shows the white-point measurement and the adjustment made to meet the target value.

Inevitably, this white-point adjustment will lead to a certain loss of luminance and contrast, caused by the required luminance reduction of individual elements of the RGB mix. It is therefore essential to specify the display with a sufficient luminance overhead to compensate for a loss that can easily fall anywhere between 10 and 30 percent. For this reason, it is important to pursue a dual approach:

•  The intrinsic display white point needs to be close to the target white point to minimize losses, including all tolerances of panel, color filter, LED bins, touch panel, and cover glass.
•  One hundred percent of display modules need to be adjusted at the end of the line – after device assembly.

Table 1 provides exemplary results for an individual display type and make. The price for achieving this result (e.g., for a specification of a ±0.005 white-point tolerance [x/y]) is a contrast and luminance loss, which was measured in this case at between 20 and 30 percent.

Table 1:  A white-point adjustment is shown for a sample display.

Parameter Sample display
  Before Target After
White Lv / cd/m2 1247   1020
          x (CIE 1931) 0.3000 0.3120 0.3118
          y (CIE 1931) 0.3301 0.3130 0.3130
          Contrast ratio 1217:1   995:1

Measuring a population of >100,000 displays of the same model acquired for production resulted in the white-point scatter in Fig. 3, which clearly shows that the intrinsic white point of this make of display falls outside the target tolerance and that the population represents a cloud of measurement points in itself. This result is typical according to the experience of the authors and is best practice for in-plane switching (IPS) LCDs with a typical automotive backlight.

Fig. 3:  This white-point scatter represents >100,000 displays from one supplier. The target tolerance field lies between the vertical and horizontal lines on the right.

Figure 4 illustrates the effect of the end-of-line adjustment algorithms for the x-coordinate (CIE 1931) of the same display.

It could be concluded that the end-of-line white-point adjustment is a mature process that meets the tolerance-window requirements. However, the luminance and contrast losses suggest that advances in intrinsic display white-point matching by display manufacturers themselves might fruitfully be made.

Fig. 4:  The intrinsic display white point on the CIE 1931 x-axis is shown in the left figure, and the matched white point of the same display population after end-of-line adjustment is shown in the right figure.

Color Adjustment

Display color adjustment is in principle an expansion of the white-adjustment process. Although more complex, the underlying approach is similar. Figure 5 shows the color-gamut triangle of two LCD panels and the target color gamut to be achieved via the adjustment process.

Fig. 5:  This CIE 1931 diagram includes the color gamuts of two LCD panels and the target gamut (light-colored dots) to be met by both.

The RGB measurement provides the basis for the adjustment that needs to be done by recalculating image data on the fly. The target of adjustment is to get the same color coordinates for red, green, blue, and gray scales on all displays. Similar to white-point adjustment, color adjustment needs to be a 100-percent end-of-line process because the tolerance chain, including display-intrinsic color differences and assembly influences such as cover/glass, touch layer, and optical bonding, encompasses so many variables that it is necessary to measure and adjust each assembled device.

The bold figures in the two right-hand columns of Table 2 show how the color- adjustment process ensures that the dominant RGB wavelengths (λdom) match the targets after adjustment. A statistical example of this adjustment process is provided in Fig. 6: the graph at the top depicts the intrinsic wavelength deviation in the red color, measured in >60,000 displays from one supplier. The effect of the adjustment process can be seen in the bottom graph of Fig. 6, in which the dominant wavelength spectrum has been calibrated to meet the target range of 623 nm (vertical line) exactly.


Table 2:  Color-adjustment process data
Parameter Sample display
  Before Target After
Red     Lv / cd/m2 237.5   230.7
            λdom/nm 620.4 623 623.1
            Sat. / % 93   89.5
Green   Lv / cd/m2 891   655
             λdom/nm 546.3 549 549.0
             Sat. / % 85   84.7
Blue      Lv / cd/m2 125   117.2
             λdom/nm 468.4 469 468.9
             Sat. / % 92   90.4

Fig. 6:  The dominant red (R) wavelength as measured in >60,000 displays before adjustment appears in the figure on the left, vs. a wavelength of the same displays after R-adjustment, as shown on the right.

The price paid for this effective color adjustment is a certain loss in saturation. Figure 7 (left) shows the saturation level for R before calibration; Fig. 7 (right) gives the same parameter after adjustment. This loss appears justified because the dominant wavelength, on the other hand, is the most important perceived parameter to the human eye – which can be quite unforgiving when confronted with color variations.

Fig. 7:  At left, the color saturation level of the >60,000 measured displays peaks at 93 percent before calibration. At right, the red color saturation of the same display shows a slight drop to around 90 percent after calibration.

Black Uniformity

Mura is an equally well-known and unwelcome phenomenon of backlit displays. Figure 8 provides an example. The prominent root causes are distortion/displacement and resulting mechanical stress that influence the panel transmission and the effect of birefringence on the display glass.

Fig. 8:  Note this example of mura caused by distortion.

To gain control over potential mura effects, Continental has defined a proprietary measurement method. This measurement jig exposes the tested display to a predefined level of displacement. Figure 9 shows the principle. One part of the display is fixed, while the opposite end is displaced by an exactly controlled mechanical force. Resulting mura is thus measured.

Fig. 9:  A proprietary jig shows the measurement method used to determine a panel’s sensitivity to stress.

By using this measurement method, it can be determined how much distortion a display may be exposed to during assembly so that the device will still meet the specification for black uniformity at the end of the line. Once this sensitivity metric is in place, it becomes possible to optimize the display and/or device design in a way that reduces mura. Analyzing many different displays in this way also aids in the understanding of root causes for mura, which in turn can help to further optimize automotive display designs in a combined and coordinated effort from both the automotive Tier 1 and the display suppliers.

The display in Fig. 10 showed an initial black uniformity of ~62 percent without displacement. Measured in the jig, a displacement of 0.6 millimeters lowered this to ~43 percent (as depicted in Fig. 10 ). This type of measurement serves to quantify just how a given panel reacts to the mechanical stress that can occur during the assembly process (e.g., time parameters, jigs), cushion-tape gluing, optical bonding, and flexible printed-circuit (FPC) assembly. To make this kind of reaction predictable, it is necessary to specify Designs of Experiment (DOE), which allow the analysis of root causes and their effects on black uniformity. Process maturity optimization can be advanced by characterizing multiple lots with a representative number of parts.

Fig. 10:  This display demonstrates a pseudocolor image of mura caused by distortion.

Key Procedures for System Integrators

The tolerance chain (meaning variations in spectral transmissions of LCD panels and spectral emissions of LEDs) of automotive displays and assembled devices necessitates adjustment procedures for white point and color. To ensure that its products meet OEM specifications, Continental has developed and implemented mature algorithms and adjustment processes. A new measurement method for black uniformity predicts sensitivity to mechanical stress. Both efforts are required to meet OEM specifications, which regularly push the boundaries of what even the latest in automotive-grade display technology can offer. To reduce adjustment-based losses, it is necessary to limit the scatter in the tolerance chain. In the case of black uniformity, it is important to develop Designs of Experiment (DOE) that help to narrow down the many factors that contribute to mura. On a system level, it remains necessary to evaluate all factors that are beyond the influence of the display supplier, which is a core task of the system integrator.  •


aIHS: Automotive Display Market Tracker Q1 2017. Page 19, Instrument Cluster Displays Shipment Share.
bFounded in 1871, Continental most recently generated annual sales of €44 billion and currently employs more than 238,000 people in 61 countries. The vehicle product portfolio for the company’s Interior Division includes: instrument clusters, multifunctional and head-up displays, control units, access control and tire-information systems, radios, infotainment systems, input devices, control panels, climate-control units, software, cockpits, and services and solutions for telematics and intelligent transportation systems. The Interior Division employs more than 46,000 people worldwide and generated sales of €9.3 billion in 2017.

Kai Hohmann and Markus Weber are with Continental Automotive GmbH in Babenhausen, Germany. Hohmann can be reached at