Camera Hardware & Software - The Huawei Mate 9 Review

Camera Hardware & Software

Huawei introduced its first dual rear camera setup with the P9 last year. Since then, Huawei has added the feature to several more of its phones, including the Honor 8 and Honor 6X. It’s no surprise then to see Huawei’s flagship Mate 9 sporting dual cameras around back.

Like the P9, the Mate 9’s rear cameras were co-developed with Leica, the well-known German camera manufacturer. The two companies worked together, tweaking lens design and Huawei’s image processing algorithms, to ensure the captured images met Leica’s requirements. The Mate 9’s Camera app also received some input from Leica. While it offers the same features as the camera included with EMUI 4.1 on the Honor 8, it adds three film modes—Standard, Vivid, Smooth—that produce the same effects as those included in Leica’s cameras. The app also uses Leica’s fonts and colors for the UI and even replicates the Leica shutter sound.

The Mate 9’s camera module uses two different sensors. For the color sensor, Huawei uses a 12MP IMX286 Exmor RS, which belongs to Sony’s stacked BSI family. This is the same color sensor used in the P9 and Honor 8. The 20MP monochrome sensor is also made by Sony but was customized for Huawei. We do not know the exact model, but based on my calculations it most likely is one of Sony’s latest sensors with 1.0µm pixels. This black-and-white-only sensor captures significantly more light than the color sensor, because it does not have an RGB Bayer color filter array keeping light from reaching the photodiodes.

By combining the luminance data from the monochrome sensor with the color data from the other sensor, the Mate 9 produces a single image that should have better contrast and less noise than an image taken with only a single color sensor. Using two smaller sensors instead of a single large sensor to capture more light also keeps the phone thinner and avoids a big camera bump; a larger sensor requires a longer focal length, adding to the z-height of the camera module.

Unlike Huawei’s first-generation dual camera in the P9, which used two sensors of the same size to achieve a one-to-one mapping between pixels, the Mate 9’s second-generation dual camera uses different resolution sensors. This implies it must scale the output from one sensor to align with the pixels of the other. To see if this extra scaling impacts image quality, I compared a series of pictures at both 12MP and 20MP. The 20MP images look sharper, like a higher-resolution image should, but otherwise I could not see any significant differences in quality. The Mate 9 still does not appear to use data from the monochrome camera to improve image quality when shooting video, but it does use it when taking HDR photos.

The improved dual-core 14-bit image signal processor (ISP) inside the Mate 9’s Kirin 960 SoC is responsible for merging data from the two sensors. The biggest change here is the integration of a second ISP dedicated to calculating distance and creating depth maps. With Kirin 950/955, this ISP, which was developed with an undisclosed industry partner, was a separate component. Bringing it inside the SoC and onto its 16nm FinFET process should reduce power consumption.

The Mate 9 uses depth sensing for longer-range focusing, part of a hybrid autofocus system that also includes PDAF (fast, works well in good lighting), laser AF (good in low-light conditions at ranges up to 2 meters or 6.5 feet), and traditional contrast AF. Huawei says all four AF methods are used in parallel in order to achieve the fastest AF performance for any condition. Only the color sensor includes PDAF, so when taking black-and-white pictures, only the depth, laser, and contrast AF methods are available. Still, this system is an upgrade over the P9 and Honor 8, whose hybrid AF system did not include PDAF at all. The Mate 9 also includes optical image stabilization (OIS), a feature Huawei’s first-generation dual camera module lacked.

The Mate 9 also uses the depth data captured from its dual cameras for its simulated wide aperture feature. This allows you to change depth of field and the subject in focus either before or even after taking a picture, useful for creating a bokeh effect. The video above shows how this works using the Honor 8 (the Mate 9 works the same way). The focus point (or point of emphasis) is selected by tapping on the desired object in the image. Sliding a finger up or down on the screen adjusts the simulated aperture between f/0.95 to f/16, which determines how much of the image appears out of focus. After applying and saving the changes, the original image and depth map are retained for making additional changes in the future.

This feature works pretty well, but it works best when the subjects of the image are within 2 meters (6.5 feet) of the camera. There are some limitations, however. This feature cannot fix a blurry, unfocused image—you need to take a good picture to start with—and it cannot be used for focus stacking. Also, it’s better at blurring backgrounds than foregrounds, especially for objects farther from the camera.

ncG1vNJzZmivp6x7orrAp5utnZOde6S7zGiqoaenZH5xhJZqZq2glWK1tq3WnqBmpZGpsm6FjKucr6GVrHx2