A third of 2022 has passed, and "high-end" is still the key word in this year's smartphone market. On April 22, the honor Magic4 to the best version of the first sale, but also triggered the industry's discussion on high-end smartphones.
According to Counterpoint, global sales of high-end smartphones grew by a record 24 percent in 2021, while the share of high-end phones reached an all-time high of 27 percent.
By this year, folding screens, self-developed chips, and higher and higher prices have repeatedly emphasized the determination of manufacturers to "high-end". Each family is crazy about the functional experience of mobile phones, and the most serious of them is undoubtedly "image".
iPhone 13 series, Huawei P50 series, OPPO Find X5 series, just premiered Honor Magic4 to the best edition, etc., the major flagships are not to video as the core selling point. The high-end battle of mobile phone manufacturers seems to have evolved into a big competition for image strength.
Even this year, everyone has gathered to engage in self-developed chips, and they have also taken ISPs as the entry point, which shows the importance of "image" for high-end.
However, how can mobile phone images be rolled up to truly dominate the high-end? That is what must be considered.
Mobile phone imaging has evolved over the past decade, and the hardware stacking is coming to an end
Smartphone imaging has evolved over the course of more than a decade, but most of the time it's been going back and forth around hardware.
Around 2011, it was the stage of smart phone pixels 1.0, from 5 million pixels to 8 million pixels to 12 million pixels, when consumers' only concept of mobile phone images was pixels.
After spelling out the pixels, you are spelling the camera. The 2014 Honor 6 Plus may not be the first dual-camera mobile phone, but it is a product that really kicks off the trend of mobile phone multi-camera, and realizes the bokeh of large aperture portraits through dual cameras, opening up a new world for mobile phone photography.
At the same time as the trend of multi-camera spreads, mobile phone photography is also igniting a war in the direction of outsole and telephoto. Sensors range from 1/2.55-inch of the iPhone 11 Pro Max to 1/1.12-inch of the Xiaomi Mi 11 Ultra, getting bigger and bigger, and zooming in at most can be more than 100 times, which is staggering.
Then, mobile phone photography entered the era of pixel piecing 2.0, 48 million pixels, 64 million pixels and even spelled to 100 million pixels. Since then, ultra-high pixels have become a new support for the "image power" of mobile phones.
It can be seen that the imaging technology of mobile phones for more than a decade has been repeatedly squeezed around the core hardware of the light perception system. So much so that today, there is little room for innovation in hardware.
As a result, people naturally shift their attention from hardware to software. The "algorithm", which has rarely been mentioned in the past, has become a savior, and it is a key weapon for future manufacturers to dominate the field of high-end mobile phone imaging.
Computational photography was born
The application of "algorithms", there is a well-known trend, that is, computational photography.
Computational photography is the integration of computer software algorithms and modern sensors, optical systems and other technologies to form a new type of photographic equipment.
It breaks through the limitations of traditional digital photography and allows algorithms to play a central role in photography, of which the application of AI technology is the key.
The first to play mobile phone computing photography was Google. On their 2016 Pixel mobile phone, they used a kind of "magic" to make the IMX378 sensor show an imaging effect that far exceeded its own performance, shocking the industry in one fell swoop.
Because IMX378 was used by many mobile phones at that time, the photos taken by the Pixel made many people exclaim, is this really the IMX378 I know?
Google's magic is computational photography. Spearheaded by Marc Levoy, the chief engineer of Google Camera and a professor of computer graphics at Stanford University, the technology is now a medicine to break the ceiling of mobile phone photography.
Honor and Apple are also the first mobile phone manufacturers to launch computational photography solutions, which are in a leading position in computational photography technology, and also drive mainstream manufacturers in the industry into the era of computational photography. In particular, the multi-main camera computational photography scheme promoted by Honor is more excellent, and the highest score of DXOMARK 146 points from honor magic4 to the best version with this core ability is enough to prove.
Honor Magic4 to the best version of the use of a comprehensive upgrade of multi-main camera fusion computing photography technology, greatly release the maximum potential of each camera, in the past when taking pictures of multiple cameras are each doing their own, the idea of glory is to change the situation of single combat, a photo can call the ability of multiple cameras, through the combination of advantages of different cameras, the picture quality, color, clarity, etc. for secondary improvement, the real meaning of the full lens participation, full focal length integration of the next generation of future images.
For example, when shooting with the main camera, the 50 million pixel color lens is integrated with the 50 million pixel spectral enhancement lens, because the spectral enhancement lens can bring more image information, so under the fusion, the clarity and sensitivity of the photo can be greatly enhanced.
For example, when shooting ultra-wide-angle photos, the ultra-wide-angle lens may be inferior to the main camera in terms of tolerance and detail processing, so that the ability of the main camera is integrated, so that the sharpness of the central area can be increased by up to 65%.
In the same way, wide-angle and telephoto lenses can be combined, and the maximum center area sharpness improvement can reach 160% on the 2x to 3.5x focal length.
This operation is a typical case of reconstructing the image process with an algorithm to break the limitation of a single hardware capability.
How good computational photography is achieved
Computational photography is so important and so strong, how can we achieve a leading position in the field of computational photography?
From the previous introduction, it is not difficult to conclude that excellent computational photography must be done to the extreme in the following three aspects:
Excellent hardware;
Self-developed algorithms;
AI performance.
Some of the top players in the industry have long regarded "computational photography" as a core weapon to enhance the competitiveness of images, such as Apple and Huawei.
Their technical paths in computational photography are different, but in the final analysis, they must do all three aspects well.
For example, the iPhone, Apple put computational photography on the iPhone XS series for the first time with the concept of SmartHDR on the table, with the help of the A12 bionic chip's powerful AI performance and powerful multi-frame fusion algorithm, so that the iPhone XS series of photography in the dynamic range far exceeded the product. Later they also developed a series of computational photography capabilities such as DeepFusion.
Huawei P50 series of computational optics is another path, the previous computational photography is mainly the idea of multi-frame fusion, and the P50 series of computational optics will be computational photography in the application of image link one step ahead, from the sensor to collect the original information this step began to integrate into the algorithm assistance, and then let the mobile phone image can collect more and better original image information.
Therefore, the photos and videos taken by Huawei's P50 series can obtain picture information and resolution power beyond its own hardware capabilities, which is also a landmark achievement.
As the ceiling of Android image flagship, the Glory Magic4 Ultimate Edition is more obvious in the "three axes" of computational photography.
As we all know, the Honor Magic4 to the best edition has a top-level configuration of the lens system, but it is not only the hardware configuration that can shake off the same competitors; in AI, it not only sits on the 7th generation AI Engine of snapdragon 8 Gen 1, but also has an independent image chip AI ISP, with an AI computing power of 28.8TOPS, combined with AI image algorithms, it is easy to achieve the preview and shooting of 4K night video.
In terms of algorithms, Honor has its own HONOR Image Engine, which can be heterogeneous with different chip platforms, isolate the strategy engine and reconstruct the image process, this algorithm is simply specially for "multi-main camera fusion", which can realize the full release of software and hardware capabilities, thereby bringing more realistic and more perfect imaging quality.
In other words, Honor's set of computational photography includes both the enhancement of information collection capabilities and the ability to fuse multi-frame images. It can be said that it is a collection of strengths.
Therefore, only by achieving the ultimate in hardware, algorithms and AI can the power of computational photography be truly exerted, and at the same time, it will also grasp the key to competing in the field of high-end mobile phone imaging at this stage.
Computational photography, how to drive the mobile phone image experience to improve the overall situation
In the competition for image power in the high-end market, computational photography is the key, but not all. After all, mobile photography is a systematic project, and every experience must be top-notch in order to truly dominate. Of course, many specific experiences are closely related to computational photography.
Today's mobile phone photography is actually a deep integration of software and hardware.
The iPhone has always been praised for the seamless switching of the multi-camera full focal length, making people feel that three cameras are working, and the immersive zoom experience is fascinating, which is the perfect combination of image algorithm and camera hardware.
The imaging capability of honor magic4 ultimate edition is not only the leading edge of computational photography, but also the leader of user innovation experience. From the perspective of sensory experience, what users need is a film that can be seen far, can be seen clearly, the night shot is clearer, and the character landscape is real and natural. Honor Magic4 Ultimate Edition is based on user experience pain points and innovation, rather than innovation for innovation, showing the ultimate rationality from the user experience.
For example, in order to create the strongest night shooting capability, honor Magic4 to the best version adopts the industry's leading 1/1.12-inch 50 million pixel super low-bottom sensor, plus independent image chip, AI image engine, bringing better signal-to-noise ratio, higher dynamic range performance, achieving super light sensitivity, and achieving brighter and clearer night scene shooting.
In the era of video and live broadcast products, video creation and Vlog have become a new trend for young people, and the Honor Magic4 to The Best Edition has launched the 4K 60fps 10bit-log movie mode, which introduces professional movie Log video to smartphones, allowing users to shoot high-definition, smooth movie-level pictures at any time. This is also an experiential benchmark for professional video creation.
From taking pictures to videos, from night scenes to portraits, from the moment when the shutter is pressed to the later stage of the image, only in every link of the photography link to achieve the ultimate experience, starting from the user experience pain points, can we truly make "image power" a winning weapon on the high-end battlefield.
epilogue
Looking back at the question at the beginning, what is the secret of mobile phone image domination? After the above discussion, Xiaobian can answer in one sentence at this moment:
Based on excellent hardware, computational photography is the core driving force, driving the full link upgrade of mobile phone photography experience.
This means that the era of "hardware is king" of mobile phone imaging has passed, and "algorithm" will become the core upgrade point. But behind the real test of manufacturers is the ability to integrate software and hardware, all from the user experience.
Whether it is the iPhone, or the Huawei P50 Pro, or the Honor Magic4 to the best version, etc., the top products in the field of mobile phone imaging are actually telling this story to users.
Image achievements are high-end, computing determines images, a new era, has quietly come.