OM System Eyes AI To Boost Camera Resolution
OM System has carved a strong identity as the go-to camera brand for travel and adventure, yet it also caters to a dedicated base of bird and wildlife photographers. The company acknowledges a key area for improvement: the relatively low resolution of its Micro Four Thirds sensors compared to the competition. To address this, OM System is considering a modern, software-driven solution.
Eiji Shirota, Head of the OM System Business Unit, spoke with PetaPixel about the company's future direction. He emphasized the power of their computational photography tools but also recognized that some users, particularly in the birding community, are hungry for more megapixels.
“With the OM-1 series, we want to be the number one camera brand for birding. But with that camera and the rest of our line, we also believe our cameras are great tools for landscape and macro, because we have so many great functions, thanks to computational photography,” Shirota stated.
The Resolution vs Speed Dilemma
For OM System, simply increasing sensor resolution isn't a straightforward fix. In March, Kazuhiro Togashi, the company's Vice President of Brand Strategy, explained the core challenge. The brand's acclaimed computational features depend on extremely fast sensor readout and processing speeds. A higher megapixel sensor would likely slow this process down, degrading the performance of features that define the user experience.
As Togashi noted, “If we increase the megapixels, we need our engine to be much stronger and much quicker to keep up with the computational photography functions... It’s hard, but it’s a good challenge.”
Shirota echoed this sentiment, reinforcing that the trade-off remains a central consideration for their development team.
A Potential Solution in AI Upscaling
Rather than waiting for a hardware breakthrough that solves the speed issue, OM System is looking at a different path: AI. The company is actively developing features like Live GND to enhance landscape photography and aims to create a shooting experience that minimizes the need for extra hardware accessories.
Shirota revealed that the team is now exploring a novel approach to the resolution problem.
“We are thinking about the improvement of the readout speed and processing, but are also considering something that combines software [and current sensor technology] with AI development, which can provide a huge jump up in resolution from the original images,” he explained.
This strategy involves integrating AI-based upscaling technology, similar to what's found in software like Adobe Photoshop or DxO, directly into the camera. The goal is to intelligently enhance images, increasing resolution and cleaning up noise without requiring a new sensor.
Navigating a Path Paved by Others
This concept isn't entirely new. Canon recently introduced a similar feature in its EOS R5 Mark II and the flagship EOS R1 to show that their 24-megapixel sensors could produce larger files. However, this implementation was met with a tepid response from the photography community. OM System will need to deliver a more compelling and effective version if it wants to win over photographers who traditionally prioritize native sensor resolution.
