Yole Développement recently published the report Status of the CMOS Image Sensor (CIS) Industry 2016: New Market and Technology Dynamics. Pierre Cambou, one of the authors, shares his vision on the evolution of the CIS industry in the smartphone market, and what Apple might do in its next iPhone.
Huawei made a big splash last month in releasing a great camera phone, the P9, which integrates a key innovation: the dual lens setup. The technical option chosen by Huawei is totally different from Apple and Samsung, but it’s competing on the same use: photography capability. The best micro cameras from Apple and Samsung are leveraging Sony’s leading CIS and the best voice coil motor (VCM) technology integrated in Autofocus (AF) and Optical Image Stabilization (OIS). Huawei’s sensor is an older Sony chip, and the VCM technology is not state-of-the-art, as it has no OIS. But doubling the number of cameras and investing in a dedicated processing unit moves it ahead of the competition. Unfortunately this comes at a price. Apple spent in the order of $17 for cameras in the iPhone 6S Plus, and Samsung reached $20 on its S7. It is likely Huawei spent close to $24 per phone for its P9 proposition. Was it worth it? Certainly, because it shows Huawei’s strength. Will Apple and Samsung follow suit? At Yole Développement we do not think so. Apple has better things to do, and at this point in time Samsung is not ready to deviate from following in Apple’s footsteps.
Mastering dual lens technology is in fact mastering the manipulation and processing of images in real time. This is only possible with a sense of depth. Yes, we are actually discussing 3D! While nobody has wanted to hear about 3D since the demise of 3D TVs and 3D gesture controllers such as Kinect or Leap Motion, 3D imaging has in fact made a comeback without us knowing. The autofocus on the iPhone 6 was controlled by special pixels called PDAF, short for Phase Difference Auto Focus. This technique already perceives the depth of field in the image. The “full PDAF” feature of the Samsung S7 is a continuation of this trend. Thanks to a double photodiode setup in every pixel, each pixel can now perceive subject distance. Huawei, by contrast, uses a system-level triangulation approach between its sensors to create a depth map of the scene. In the end, the 3D information is extremely valuable for the primary usage of smartphone cameras: photography. In Huawei’s case it also enables a clean “camera bump”- free smartphone. In our view Apple could use 3D sensing technology for something of greater value.
(Source: Status of the CMOS Image Sensor (CIS) Industry 2016: New Market and Technology Dynamics, Yole Développement, June 2016)
The takehome message is that Apple will use 3D sensing technology in its front facing cameras, either with a dual lens setup or with an additional 3D image sensor. It has been in Apple’s DNA for 35 years now to innovate on the Human to Machine Interface (HMI) side of computing. There is no way the recent advances in this field have not been fully taken into account by the famous Cupertino-based company. Its recent purchases, Linx, Metaio and more recently Faceshift, show that they have been working on the generation and usage of 3D imaging technology for a few years. From our point of view at Yole, the timing is right in terms of technology. It is also right for Apple to boost its sales with an innovative offering. We think that by leveraging 3D sensing technology it is now possible to transform the live video call experience of Facetime and other Skype lookalikes. Being able to remove the background, create avatars, to look into the eyes of the caller while eventually using a VR headset has the potential to be truly transformative. Apple is best placed to get it right. Welcome to the third dimension.
Authors: Pierre Cambou & Guillaume Girardin from Yole Développement
New functions are pushing change in CMOS image sensors, boosting the market toward $18.8B in 2021 at 10.4% CAGR
Related Reports and Monitors
Triple Forward Camera from Tesla Model 3
Reverse Costing - Structural, Process & Cost Report
Artificial Intelligence for Medical Imaging 2020
Market & Technology