1. Semiconductor Wafer
A semiconductor wafer is a thin slice of semiconductor material used as the substrate for manufacturing integrated circuits and image sensors. Silicon is the most widely used material in semiconductor wafer production.
The manufacturing process begins with high-purity polysilicon, which is melted in a controlled environment. A seed crystal is introduced into the molten silicon, and a single-crystal silicon ingot is gradually grown. The ingot is then ground, sliced into thin wafers, and polished to achieve the required surface quality.
Wafer diameters of 8 inches (200 mm) and 12 inches (300 mm) are widely used in semiconductor manufacturing. The wafer provides the foundation on which the sensor's electronic circuits and pixel structures are fabricated.
2. Bayer Color Filter Array
A Bayer filter is a color filter array (CFA) commonly used in CMOS image sensors to capture color information. Its standard RGGB pattern consists of 50% green filters, 25% red filters, and 25% blue filters.
Individual image sensor pixels measure light intensity rather than directly identifying full-color information. The Bayer filter allows different pixels to capture different color components. An image signal processor (ISP) then reconstructs a full-color image through a process known as demosaicing.
One limitation of the Bayer pattern is its susceptibility to color artifacts when capturing repetitive fine details, such as fabric patterns, grids, or closely spaced lines. These artifacts can appear as false colors or moiré patterns because the scene contains spatial frequencies that interact with the sensor's pixel grid.
An optical low-pass filter (OLPF), sometimes called an anti-aliasing (AA) filter, can reduce moiré and aliasing by slightly softening high-frequency image details before they reach the sensor. However, this reduction in fine detail may also decrease perceived image sharpness.
The choice of whether to use an AA filter depends on the application. Systems that prioritize maximum detail may omit it, while applications that frequently encounter repetitive patterns may benefit from optical low-pass filtering or suitable image processing.
3. CMOS Image Sensor
CMOS stands for complementary metal-oxide-semiconductor, a semiconductor technology widely used to manufacture integrated circuits. CMOS image sensors use this technology to convert incoming light into electrical signals that can be processed into digital images.
CMOS technology has become increasingly important in digital imaging because it offers several advantages over traditional charge-coupled device (CCD) technology, including flexible readout architectures, high integration potential, low power consumption, high-speed operation, and the ability to integrate additional functions on the same chip.
Depending on the sensor design, these advantages can support higher resolution, improved sensitivity, wider dynamic range, faster image capture, and more compact system integration.
A typical CMOS image sensor package or pixel structure may incorporate several functional layers, including:
Microlens layer: Directs incoming light toward the photosensitive regions.
Color filter array: Separates incoming light into red, green, and blue components.
Wiring and interconnect layers: Carry electrical signals and connect pixel circuitry.
Photosensitive layer: Converts photons into electrical charge.
Silicon substrate: Provides the semiconductor foundation for the device.
The exact layer arrangement varies by sensor architecture. In front-side illuminated (FSI) sensors, wiring is located above the photodiodes. In back-side illuminated (BSI) sensors, light enters from the opposite side, reducing the obstruction caused by front-side wiring.
The Role of Microlenses and CRA
Because light may reach individual pixels at different angles, microlenses are positioned above the pixels to direct as much incoming light as possible toward the photodiodes.
The effectiveness of this optical arrangement depends partly on the chief ray angle (CRA) of the lens and the sensor's microlens design. These parameters should be matched during camera module development to maintain optical efficiency across the image.
4. Chief Ray Angle (CRA)
What Is CRA?
The chief ray angle (CRA) describes the angle between the chief ray passing through the optical system and the optical axis, evaluated at a specified location, typically the image sensor plane.
In camera module design, the lens CRA describes how the chief ray reaches different positions on the sensor. The sensor's CRA specification indicates the angular response that its microlens and pixel structure are designed to accommodate.
These two parameters are related but are not identical. The lens CRA describes the incident light geometry, while the sensor CRA specification describes the sensor's intended angular response or microlens configuration, depending on the manufacturer's definition.
The lens and sensor CRA profiles should therefore be evaluated together rather than treating either value in isolation.
Why Is CRA Matching Important?
The angle at which light reaches a pixel affects how efficiently that light is collected. Rays near the optical axis generally approach the sensor at relatively small angles, while rays toward the edges of the image may arrive at larger angles.
If the lens CRA profile is poorly matched to the sensor's microlens arrangement, some incoming light may not reach the photodiodes efficiently. This can contribute to corner shading, reduced peripheral sensitivity, color shading, or uneven image brightness.
As camera modules become more compact and lenses become shorter, managing the incident angle of light across the sensor becomes increasingly important.
A commonly used preliminary design guideline is to keep the difference between the lens CRA and the sensor's specified CRA within approximately ±2° to ±3° where applicable. However, this is not a universal acceptance criterion. The appropriate tolerance depends on the sensor manufacturer's specifications, lens design, image height, wavelength, and the intended application.
A smaller lens CRA is not automatically better. The objective is to achieve compatibility between the lens's angular output and the sensor's optical response across the entire image field.
CRA and Vignetting
Several factors can contribute to shading or vignetting in a camera system.
One is the angular response of the sensor's microlenses and pixel structure. Another is optical vignetting, which occurs when the lens barrel or other optical elements physically restrict the light reaching the sensor.
The cosine-fourth-power law (cos⁴ law) describes one component of natural illumination falloff in an idealized optical system. Actual shading also depends on lens construction, pupil geometry, sensor response, and mechanical obstruction.
Sensor microlens optimization can improve light collection at larger incident angles. Image signal processors also commonly use lens shading correction (LSC) to compensate for residual brightness and, in some systems, color nonuniformity across the image.
LSC can improve image uniformity, but it cannot fully recover information lost through severe optical obstruction or compensate for every source of optical degradation.
5. How CRA Affects FSI and BSI Sensors
Front-Side Illuminated (FSI) Sensors
In a conventional FSI sensor, incoming light passes through the microlens and color filter array before reaching the photodiode. Metal wiring and other structures are located above the photosensitive region and can obstruct some incoming light.
When the incident angle increases, more light may be blocked or redirected away from the photodiode, depending on the pixel layout and microlens design. This can reduce light collection efficiency, particularly toward the edges of the image.
Larger pixels can sometimes reduce the relative impact of this obstruction because their photosensitive regions and optical structures may provide greater tolerance for off-axis rays. The actual benefit depends on the specific sensor architecture.
Back-Side Illuminated (BSI) Sensors
BSI sensors are designed so that light enters through the back of the silicon substrate rather than passing through the front-side metal wiring first.
This arrangement reduces the obstruction caused by front-side interconnects and can improve light collection efficiency, especially in small-pixel sensors or applications with challenging illumination angles.
BSI technology does not eliminate all CRA-related limitations. Microlens design, pixel isolation, sensor thickness, optical stack construction, and lens characteristics still influence the sensor's angular response.
6. How to Select a Sensor CRA for Different Lenses
Selecting a sensor for a camera module requires consideration of the lens's CRA profile, the sensor's angular response, the field of view, and the intended operating conditions.
Wide-Angle Lenses
Wide-angle lenses often produce larger chief ray angles toward the edges of the image. The sensor should be selected with an angular response compatible with the lens's CRA profile.
Sensors designed to accommodate larger incident angles, including suitable BSI sensors, may be beneficial in applications such as smartphones, security cameras, toys, and webcams.
A CRA specification above 25° may be worth evaluating for certain designs, but it should not be treated as a universal requirement for all wide-angle lenses. The actual lens and sensor specifications must be checked together.
Telephoto Lenses
Long-focal-length lenses may have relatively small CRA values, depending on their optical design. In these systems, a sensor with a compatible low-CRA response may be appropriate.
Applications can include security monitoring and machine vision. A nominal sensor CRA of 0° should not be assumed necessary for every telephoto lens; compatibility must be confirmed using the manufacturer's specifications and optical design data.
Zoom Lenses
The chief ray angle may vary with focal length, zoom position, focus distance, and image height. Sensor selection should therefore consider the complete CRA range rather than a single nominal value.
Sensors with larger pixels or BSI architecture may offer advantages in some designs, but neither characteristic guarantees better performance in every zoom system.
Optical simulation, CRA data comparison, and prototype testing across the operating range are recommended when developing zoom camera modules.
7. Practical Recommendations for Camera Module Design
When selecting a CMOS image sensor and lens combination, engineers should evaluate the following factors together:
Sensor specifications: Pixel size, sensor format, sensitivity, CRA definition, and angular response.
Lens specifications: CRA profile, focal length, aperture, field of view, and relative illumination.
Sensor architecture: FSI or BSI structure, microlens arrangement, and pixel design.
Image quality requirements: Resolution, color accuracy, corner brightness, and low-light performance.
ISP capabilities: Lens shading correction, color shading correction where supported, and other image-processing functions.
Validation: Test images across the full field of view and under representative lighting conditions.
A well-matched lens and sensor combination can improve light collection efficiency, reduce shading, and deliver more consistent image quality. For OEM and ODM camera module projects, CRA compatibility should be evaluated early in the optical design process rather than treated as a final-stage adjustment.
