Huawei on their cameras: “quality over megapixels”

In an interview with, Huawei explains their camera strategy for example with the P40 Pro or the P40 Pro Plus. This way they want to maintain their status as the camera flagship by saying: “Quality over Megapixels”.

Smartphone cameras in general.

“A camera can be compared to a watermill, because to make it perform better you can make several adjustments such as increasing the wheel and adjusting the size of the tubs, but the essence remains to use the flow of water to create energy.” explains Pan Chao Yue, a camera expert who works at Huawei’s Shanghai facility. “A camera is all about light instead of water to take good pictures and you can use a larger sensor size to let the camera catch more light or adjust the color filter array,” says Pan Chao Yue. Huawei does not opt for a classic RGB sensor (Red Green Blue) for its camera, but designed its own RYYB variant (Red, Yellow, Blue) that captures more light.

Pan Chao Yue explains that with the P40 series, Huawei has added the largest camera sensor you can find in a phone, and it has a size of 1 / 1.28. “Of course, the size of the sensor goes together with the place that you can free up and that can cause limitations,” explains the camera expert. Every manufacturer puts different stresses in the design when developing phones, and the camera is an important aspect for the P40 series, according to the Huawei expert.

Computational Photography.

Another aspect that Huawei would like to emphasize when it comes to camera development is the importance of software. “Computational photography is something that we are learning more and more and it is also becoming more important. You could say that with our smartphone cameras now the hardware and software both contribute 50 percent to the performance, and they go hand in hand,” explains Pan Chao Yue out. An example where camera hardware and software come together is with ‘AIS’, which is the name that the Chinese brand gives to the software improvement of optical image stabilization with its cameras. Huawei uses machine learning models to assist the optical stabilization of its camera.

Huawei’s P40 Pro Camera

Main Camera

The primary camera comes with a very large 1/1.28″ 50MP Quad-Bayer sensor that produces 12MP image output. As with other devices with similar sensors, it uses pixel binning to increase dynamic range and low-light capabilities. The lens features an f/1.9 aperture and nominal 23mm-equivalent focal length; however, Huawei crops the field of view to a more conventional 27mm. This decision was presumably made in order to design a thinner camera unit, but it should also help with reducing such artifacts as corner softness and distortion. Optical image stabilization, which can be challenging to implement on such large image sensors, is on board as well.

Ultra-wide camera

Like with the Mate 30 Pro, Huawei implements a large and high-resolution image sensor (40MP 1/1.54″) in its ultra-wide camera, which combined with a fast f/1.8 lens, should make for excellent low-light image quality. The downside of such a large sensor in the ultra-wide is lens design. It’s a challenge to design a lens that provides a very wide angle of view and still fits into the thin body of a high-end smartphone. This is why Huawei settled on a compromise at 18mm, which is noticeably wider than the main camera, but not as wide as that of some competitors, such as the latest Samsung devices and iPhones, which offer 12mm and 13mm lenses, respectively.

“Field-of-view fusion” zoom

The P40 Pro doesn’t rely just on hardware alone for zooming, however. Like the P30 Pro, the P40 Pro uses software systems which Huawei calls field-of-view fusion and AI RAW, a combination of optical and algorithm-powered digital zooming that adapts to the chosen magnification factor. Only the primary camera is used for up to 2x magnification; the tele-camera takes over from 5x and longer. For intermediate zoom factors, the P40 Pro uses a fusion algorithm that combines image data from both cameras, merging several RAW frames from the main and tele cameras into one high-resolution frame, which is then cropped for zooming.

For optimal detail, the 5x tele-cam records image data at the center of the frame; it uses AI-refined image data from the primary camera to fill in the “missing” image areas around the edges. A deep learning algorithm improves detail on fine patterns and textures—quite an impressive feat, considering that autofocus, white balance, and other image parameters of both cameras all have to be in perfect sync for good results.

At the long end, the camera is capable of achieving a 50x zoom factor using a combination of optical zoom and super-resolution processing that stacks multiple RAW frames from the tele-cam.

Lämna ett svar

Din e-postadress kommer inte publiceras. Obligatoriska fält är märkta *