After the release of Snapdragon's next-generation flagship mobile platform, the third-generation Snapdragon 8, in October, the mobile phone market has launched a large number of new products in the past month or so. It is obvious that a trend can be seen: in the recent wave of new Android products, in addition to the regular performance upgrades, the image has once again become the focus of the competition.
In previous years, the annual image improvement was mainly reflected in some large and super large cup models, but this year, the standard models of mainstream brands have also comprehensively strengthened their imaging capabilities. The further involution of manufacturers in mobile phone imaging has brought more benefits to consumers, and has also promoted the iteration and upgrading of imaging capabilities in the entire industry.
The standard version of the image begins to roll in.
There are many products that can reflect this comprehensive upgrade, and the GT5 Pro released by Realme is a typical representative of it. This third-generation Snapdragon 8 flagship, priced at 3399 yuan, even surpasses many more expensive flagships at the image level. Sony's new generation of LYT-808 main camera scheme + IMX355 ultra-wide angle + IMX890 periscope telephoto, this hardware scheme can be called luxurious.
The Realme GT5 Pro is not only equipped with a high-spec main camera, but also adopts a periscope telephoto solution that has rarely been seen before in its class: IMX890. You know, the IMX890 was previously used as the main camera solution for some models in the same price range, and now it is directly used as a telephoto solution by the realme GT5 Pro.
Also reflecting this year's comprehensive image upgrade models are Xiaomi Mi 14, on this standard version of the small screen flagship, Xiaomi is equipped with a large bottom light and shadow hunter 9000 main camera, with 75mm floating telephoto and 50 million pixels of ultra-wide angle scheme. In terms of image configuration, the Xiaomi Mi 14, priced at 3999 yuan, is very close to its own super-large flagship Xiaomi Mi 14 Pro. Users can experience the full Leica image on this standard version of the small-screen flagship.
The ability to deploy a powerful imaging system on a standard model is inseparable from the imaging support provided by the mobile platform. In particular, the current camera flagships all use ultra-high pixel multi-camera solutions, which requires the mobile platform to provide sufficient ISP performance to provide underlying support for the multi-camera system. In fact, the Snapdragon mobile platform has always been "ahead of the curve" in this regard: as early as the Snapdragon 888 era, the Snapdragon platform supported three 14-bit CV-ISPs, and supported multiple cameras for triple concurrency and parallel processing, and the latest third-generation Snapdragon 8 platform supported three 18-bit ISPs.
Snapdragon imaging technology is everywhere.
What is the real help of powerful ISP performance for mobile phone imaging?To give you a common example, the Xiaomi Mi 14 has the "what you see is what you get" capability in high dynamic range images this time. In other words, when you use the Xiaomi Mi 14 to shoot some scenes with a large light ratio, such as night scenes and sunsets, the device outputs high dynamic range images in real time during the camera preview. This way, you'll be able to control the final image precisely before you press the shutter. Whether it is for the shooting experience or the final film, it is of great help.
This feature may seem insignificant, but it puts extremely high demands on the performance of ISPs. It requires the sensor to collect and output images at the resolution of the camera when the mobile phone takes a photo preview, and the ISP receives and processes the image data in real time. You must know that the current high-performance image flagships are all ultra-high pixel configurations, and multiple algorithms such as HDR and night scene mode have also been introduced. This means that the amount of data and computation that ISPs need to process in real time is extremely large. It is the strong support of Snapdragon Sight Snapdragon imaging technology based on the third-generation Snapdragon 8 mobile platform that enables imaging flagships like the Xiaomi Mi 14 to deliver excellent imaging performance.
In the current mobile phone market, the most involuted telephoto scheme, Snapdragon imaging technology is also ubiquitous.
Realme GT5 Pro is equipped with the industry's first third-generation Snapdragon 8 + IMX890 telephoto combination this time, which is rare in products of the same level. In fact, this scheme is the result of in-depth joint debugging between Realme and Qualcomm. This product is equipped with a supercore image heterogeneous computing framework, which supports the camera's single-frame processing speed by 44%, reduces image power consumption by 10%, and supports multi-frame synthesis of up to 13 frames. In order to improve the image quality performance of IMX890 in extreme scenes, Realme and Qualcomm have cooperated to support DOL-HDR technology for the first time in the telephoto terminal. This technology is a new multi-frame optimization strategy through 2 6+1, which can further improve the latitude and dynamic range of the picture, and the picture is more pure and realistic. In other words, similar HDR technology can improve the latitude of the picture through multi-frame compositing, but it will also distort the picture. The solution of Realme and Qualcomm can improve the image quality and dynamic range performance as much as possible on the basis of ensuring authenticity and naturalness, and greatly improve the overall look and feel of the telephoto picture.
This technology also brings a benefit, because Dol-HDR takes two different frames at the same time for each shot, even when shooting high-speed moving objects, you can still get enough picture information to synthesize high-dynamic HDR**, so as to improve the finished film. It is the cooperation between Realme and Qualcomm that finally contributed to the realme GT5 Pro's leapfrog performance in imaging capabilities.
AI imaging layout in the future.
If these examples are the embodiment of the Snapdragon platform's support for the underlying imaging capabilities, then the development of the Snapdragon mobile platform in AI imaging over the years is expected to completely rewrite the future of mobile phone imaging.
If you've followed the annual Snapdragon Tech Summit, you'll be impressed by the ubiquity of AI at the summit. From a very early stage, Qualcomm has been emphasizing the AI features of the Snapdragon mobile platform. At this year's Snapdragon Technology Summit, AI has almost become the "main theme" of the entire summit.
As early as last year's Snapdragon 8 Gen 2 mobile platform, Qualcomm brought a new design of "cognitive ISP", which can achieve automatic enhancement of **and** through real-time semantic segmentation. Image segmentation refers to the process of subdividing a digital image into multiple image subregions or superpixels. The aim is to simplify or change the representation of the image so that it is easier to understand and analyze. On the basis of understanding and analysis, the image is deeply optimized through powerful AI to further improve the look and feel.
On the third-generation Snapdragon 8 platform, semantic segmentation can support up to 12 layers, making the recognition and segmentation of objects and scenes more accurate. Even if it is an otherwise plain **, through semantic segmentation and the intervention of AI images, it can present a more wonderful and natural color effect. At the functional level, the third-generation Snapdragon 8's Super Night View supports taking photos and videos, and improves efficiency and energy efficiency through the introduction of a large number of AIs. There's also a Vlogger View mode, which allows you to use dual front and rear cameras to combine with your ISP to capture two images at the same time.
In terms of generative AI, which has been popular in the past two years, the third-generation Snapdragon 8 supports multi-modal general-purpose AI models and large models with more than 10 billion parameters. Harnessing the power of generative AI, the Snapdragon platform supports snazzy new features such as AI expansion. In the long run, the development of functions and usage scenarios based on generative AI has become an important research direction for current mobile phone terminal manufacturers, and the Snapdragon platform has made a forward-looking layout and foreshadowing in this regard.
At present, in terms of mobile phone imaging, the Android platform is already the de facto industry leader. Whether it is the existing mobile phone imaging capabilities or the development of future-oriented imaging technology, Android imaging flagships have always been at the forefront of the industry, so there is a saying that "mobile phone images look at Android". However, looking back at the development history of mobile phone imaging over the years, we can also see that the Snapdragon platform has played a key role in it, which is reflected in the underlying support capabilities provided by each generation of Snapdragon mobile platform based on the needs of mobile phone imaging. It is also reflected in the leading role provided by Snapdragon as a platform itself for the development of mobile phone imaging technology.
What's more, Snapdragon has also devolved these advanced imaging technologies to more platforms over the years. In the past two years, not only the Snapdragon 8 mobile platform, but also mid-range products, including the Snapdragon 7 series family, have also been equipped with flagship-level imaging capabilities, and the imaging capabilities of many mid-range products have also been improved by leaps and bounds. The flagship imaging capabilities that users had to spend four or five thousand or even more to obtain before can now be experienced in a more accessible way, which is also inseparable from the role of the Snapdragon platform.
In other words, the current battle of mobile phone images has entered the second half. In the past, everyone fought hardware and volume algorithms, which made great progress in mobile phone images. In the second half, each company not only needs to fight for hardware and algorithms, but also needs to fight for AI capabilities, but also for platform ecology and integration capabilities. From this point of view, if the current mobile phone image depends on Android, and the future of Android image has to look at Snapdragon.