Sony IMX908 4K Camera Sensor Unveiled

Sony Semiconductor Solutions Corporation (hereinafter: Sony) announced the upcoming commercialization of the “IMX908” sensor. This sensor is a 4K resolution CMOS image sensor employing the industry’s smallest 1.45µm LOFIC pixels, for use in security cameras. Sample shipments are scheduled to begin at the end of March 2026.

This product employs newly developed LOFIC pixels, enabling shooting with both 4K resolution and a high dynamic range of 96dB in a single exposure. Furthermore, by improving low-light performance, it achieves high-quality imaging with less overexposure, underexposure, and noise than conventional models in environments with large differences in brightness or in dark places.

Sony IMX908 Development Philosophy

In the field of security cameras, where high-precision image recognition is required in various environments both indoors and outdoors, the company is expanding its product lineup to include products that combine high resolution and high dynamic range, thereby contributing to the realization of a safe and secure society.

Security cameras are becoming widespread in society for a wide range of applications, not only for monitoring safety management but also for understanding the state of living spaces such as urban areas and facilities. Recently, AI-based image recognition has become standard equipment, and there is a growing need for image sensors that can reliably capture high-quality images in both bright and dark conditions.

This product features Sony’s proprietary LOFIC pixel technology “STARVIS 3,” newly developed for security cameras. Compared to previous models, it expands the saturation charge capacity by approximately 20 times while simultaneously improving low-light performance by approximately 27%, achieving a dynamic range of 96dB.

Furthermore, this product achieves high dynamic range in a single exposure, unlike the multiple exposures common in HDR shooting, enabling high-definition image output with reduced artifacts even in scenes containing moving objects. In addition, these characteristics are realized with the industry’s smallest 1.45µm LOFIC pixels thanks to a unique pixel design. By enabling high-quality 4K shooting in environments with large differences in brightness or in dark places, it contributes to improving the recognition accuracy and multi-functionality of security cameras.

Sony IMX908 Main Features

Industry-leading smallest 1.45µm LOFIC pixels enable both compact size and 4K resolution.

This product incorporates newly developed LOFIC pixels. The LOFIC structure enables more efficient charge storage and voltage conversion than conventional methods, contributing to an increased sensor saturation charge capacity and improved low-light performance. By realizing this structure with a single pixel measuring 1.45µm, the smallest in the industry, it enables 4K resolution shooting despite its compact 1/2.8-inch size.

Dynamic range of 96dB, enabling high-quality imaging in environments with large differences in brightness and in dark places.

This product expands the saturation charge capacity by approximately 20 times compared to conventional models, allowing it to store more charge and enabling shooting in brightly lit environments with reduced overexposure. Furthermore, because it can convert voltage from less light than before, low-light performance is improved by approximately 27%, suppressing underexposure and noise in dark conditions. These improvements expand the dynamic range achievable with a single exposure to 96dB, enabling high-sensitivity, high-quality imaging even in environments with greater contrast or in dark conditions.

The single-exposure method enables high-definition imaging with reduced artifacts.

This product achieves high dynamic range with a single exposure method. Unlike the multi-exposure method, which combines multiple images taken with different exposure times, it enables image capture with reduced artifacts and high-speed image output. As a result, it can stably capture high-definition images of moving subjects while suppressing contour shifts and color shifts that can hinder AI image recognition, thereby contributing to improved recognition accuracy.

Furthermore, this product can output multiple image data generated with different conversion efficiencies, providing more flexible options for camera design.