For years, the official Raspberry Pi Camera Module has been one of the best accessories for Raspberry Pi boards, giving camera capabilities to any DIY project. Now a new revision of the camera is available for purchase.
The Raspberry Pi Camera Module 3 was announced today, replacing the Camera Module 2 that has been around since 2016. The new accessory is available in four models, and is compatible with all Pi boards with CSI connectors, except the Raspberry Pi 400 and original Pi Zero. Even though the dimensions and mounting holes are identical compared to earlier models, the size and position of the sensor modules have changed, so it won’t be compatible with cases designed for older camera modules.
The new Camera Module 3 uses Sony’s IMX708 sensor, the same one used in the Oppo Find X2 phone from 2020. It provides a 4608×2592 image, or 11.9 MP, which is a notable improvement over the older 8 MP Camera Module 2. The upgraded sensor allowed Raspberry Pi to implement HDR support, which combines the same image at different exposures to create a more a balanced image. That feature has been available on most smartphone cameras for years, but better late than never, right?
Raspberry Pi also added powered autofocus, a completely new feature, so the lens physically moves backwards and forwards to improve focus. The company said in a blog post, “to select the appropriate lens position, we use the Phase Detection Autofocus (PDAF) capabilities of the IMX708 sensor, falling back to our own Contrast Detection Autofocus (CDAF) algorithm if a high-confidence PDAF result is not available. A nice bonus of PDAF is that it allows us to run the autofocus algorithm continuously during video recording, maintaining optimal focus as the camera, and objects in the scene, move.”
Pricing starts at $25 for the basic Camera Module 3, with the wide-angle version available for $35. Keep in mind the cameras only work with Raspberry Pi’s newer libcamera and Picamera2 software, and not the older closed-source software that was built for earlier models.
Source: Raspberry Pi