In the world of mobile imaging, there’s a long-held notion that the sensor, lens (or a combination of several) and software share equal importance. If we were to convert that to a numerical expression, it would look something like 1:1:1.
However, as technology matures, we’ve begun seeing homogenisation among the industry’s top players. Almost every recent flagship smartphone hitting the market today has a multi-camera setup, with manufacturers shifting focus to other differentiators, trying to find their own niche.
Many have ended up looking towards software, so much so that it’s shifting the balance of industry norm. While for the longest time it has been 1:1:1 – it’s now begun to look a lot more like 1:1:2, with software being represented by the “2.” From here we can see that companies have been placing emphasis on software.
Apple has its “computational photography mad science” Deep Fusion technology, Samsung most recently introduced Single Take with the Galaxy S10 series, and HUAWEI, during the launch of the HUAWEI P40 Series, introduced to the world the XD Fusion Engine. And today, we take a look at what this “engine” really is.
Huawei XD Fusion Engine
The HUAWEI XD Fusion plays a key role among many superior computational photography capabilities that the HUAWEI P Series has to offer. Before we dive into the technicalities, here’s the simplest explanation to what it does: it lets you take better photos.
It’s really that simple. It doesn’t matter whether you’re shooting in bright daylight or at night, a person or a thing you have on your desk, the XD Fusion Engine optimises for detail and image quality to give you the best final image.
But then of course, as much as we hope it is, it’s not magic. It’s the company’s engineering ingenuity, integrated into a complete solution that brings a noticeable improvement to any and all photos.
The XD Fusion Engine conducts high-level semantic analysis on multiple RAW frames captured on multiple cameras. The high-fidelity images are deconstructed into Scenarios, Biological Information, Body Mask, Skin Mask, Face Parsing and Face Landmark layers. These components are then fed to the NPU for processing.
Leveraging the Da Vinci-based NPU inside the Kirin 990 chip, XD Fusion Engine enhances the images mimicking how the human brain processes visual information. The pixel-level imaging processes detail, colour, dynamic range, bokeh and other factors to generate a high quality portrait image for the end users.
Here are some of the photo samples we took with the Huawei P40 Pro, during our time with it for the review:
Aside from the extraordinary XD Fusion Engine working in the background to optimise your photos, the HUAWEI P40 Series also includes some new tools for you to quickly edit your images.
The Huawei P40 Series also has the capability to support real-time AI recognition and processing. These allow the Huawei P40 Series to offer three new functions:
- AI Best Moment
- AI Remove Passersby
- AI Remove Reflection
These features are all collectively referred to as Huawei Golden Snap. They augment moving pictures (Moving Picture mode needs to be turned on for this to work). For AI Best Moment, the Huawei P40 Series detects and compares more than 30 posture detection and over 90 facial expressions on subjects in the frame to select the best image, and presents it to the user as a recommendation as they edit their images in the Photo Gallery. AI Remove Passersby identifies passers-by on the images for easy removal. Finally, AI Remove Reflection scans for reflections and glare and removes them from the final photo.
HUAWEI has been advancing rapidly when it comes to computational photography. From the P10’s scene recognition, to the P20 series’ Master AI and Night Mode, the P30 series’ AI Colour, and now the P40 series’ XD Fusion Engine.
We’ve tested the Huawei P40 Pro, and it’s definitely one of the best camera smartphones we have ever had on our hands. If you haven’t already, check out the full review here.