As Vivo introduces its X80 series in India, the focus like many other flagships remains on the camera capabilities. But there’s a lot going on behind the scenes with the cameras on these phones. While the X80 series continues its partnership with renowned lens and camera brand Zeiss, the brand is also putting more focus on artificial intelligence and its own dedicated V1+ plus imaging chipset. This, according to the company, plays a key role in helping improve image quality on the new range of premium devices.
“Smartphone photography is half computation and half hardware. With our V1+ imaging chip and with traditional processors getting better and faster every year, there is a lot more neural processing, AI work that is now going into imaging and it will keep getting better,” Keshav Chugh, Product Manager for Image Effects, Vivo told indianexpress.com during an interaction in Dubai. Chugh is heading the company’s research efforts into improving its camera back in Shenzhen, China.
For Vivo X80, the company is using Sony’s IMX866 for the primary 50MP camera, while the high-end Pro gets the Samsung GNV sensor for the 50MP camera. Vivo says both have been customised for their phones, and according to Chugh, the sensor size also makes a difference.
“The bigger the sensor, the more light that’s going in the camera and the more details we have to work with to compute. We are using glass lenses supplied by Zeiss. So that’s the physics part sorted. And the second factor like I said, computation is the future and this is why we developed the V1+ chip,” he said.
The V1+ can do all the heavy lifting when it comes to running the resource-intensive algorithms required for imaging processes such as a super night mode (for low-light photography) or an AI super-resolution mode or even for the AI skin feature. “There are multiple algorithms running parallel on top of each other. Running these just on the NPU and the GPU of the SoC (system on chip) is not very effective. It’s slow, very high in power consumption and the GPU is known to get become very hot,” Chugh explained. The NPU is the neural processing unit while the GPU is the graphics processing unit on a smartphone.
Keshav Chugh, Product Manager for Image Effects, Vivo.Another advantage of the V1+ is that it allows for algorithm scheduling to reduce time spent on image processing. “We can run more algorithms and do more computations in real-time. That’s why our camera effect would be much better compared to other phones with the same processor and ISP,” Chugh added.
He also pointed out that some of the Zeiss filters that they have included in the device are not a simple bokeh mode and require a lot of algorithms. The company is also introducing a new feature called AI skin on its X80 series, which aims to give a more natural look to user photos while preserving skin texture.
According to Vivo’s executive, the company has in fact spent millions of dollars trying to figure out how to implement artificial retouching. “We paid professional retouchers to retouch portrait pictures for us. They spend four to five hours on each picture. And then our team in Hangzhou developed a neural network machine learning algorithm to learn from the before and after images. So our AI can now retouch the picture by itself while maintaining the skin texture. Say if you have freckles or moles on the face, we maintain all your natural real features,” Chugh explained.
“Our direction now is going towards a very natural effect,” he added and said that they are taking into account customer feedback as well. Vivo states that in future all V and X series will have this algorithm. He added that one reason it will be limited to mid-range devices is that it is a processor-heavy algorithm to run.
The Vivo X80 series goes on sale in India starting on May 25 at a price of Rs 54,999 for the base X80, while the Pro version costs Rs 79,999.
Disclaimer: Vivo sponsored the correspondent’s flight and hotel stay in Dubai for the X80 immersion session.
