Canon announces development of the EOS R1 flagship with deep learning technology promising improved image quality
Canon has announced that the ‘flagship’ Canon EOS R1 is under development. While that’s not surprising, what is telling from the rather scant details is that AI will feature heavily.
Reading between the lines I expect to see noise reduction improved over the camera’s internal processing by the bundled DPP software, in effect locking out other parametric editors such as Capture One and Lightroom from accessing the data and ultimately the promised noise reduction improvements.
Not only is this to add value to the camera and system by leveraging the software side, but we will likely be choosing our camera systems based on software enhancements in the future. I wouldn’t be surprised to see more of this from rivals Sony and Nikon going forward.
I’ll be happy if I can control or reprogram some esoteric features, such as setting an aperture series, while tethering from an iPad or iPhone.
Press Release (auto-translated from Japan)
The “EOS R1” currently under development is a professional mirrorless camera that combines Canon’s state-of-the-art technology and combines the highest-class performance, high durability and high reliability required for flagship machines. We will dramatically evolve both still images and video performance※, and respond to the high needs of professionals who are active in the front line in a wide range of fields such as sports, news, and video production.
This machine adopts the newly developed [processing] engine “DIGIC Accelerator” in addition to the conventional [processing] engine “DIGIC X”. A new [processing] engine system, including the newly developed CMOS sensor, enables high-speed processing of large amounts of data and brings unprecedented evolution, including AF.
By fusing the new [processing] engine system and deep learning technology in high dimensions, high-speed and high-precision subject recognition is realized. For example, in team sports where the accuracy of tracking the subject is improved and the subject intersects, it is possible to keep track of the target player even if other players overlap in front of the player you want to shoot. In addition, the “action priority” function of AF recognizes the action of the subject by analyzing the situation of the subject at high speed. In sports scenes where it is difficult to predict the development, players who are doing specific actions such as shooting are automatically determined as the main subject, and by instantly moving the AF frame, it supports the shooting of the decisive moment.
The combination of the new [processing] engine system and deep learning technology also contributes to the improvement of image quality. Canon will realize image noise reduction that Canon has established in advance with PC software with the camera function, and will contribute to the user’s creative activities by further improving the quality of the video.
Canon will proceed with the field test of this machine and support the shooting of decisive moments and moving moments at future global sports events.
The “EOS R System” will continue to respond to a wide range of users’ needs and contribute to the development of photo and video culture by expanding its lineup of attractive cameras and “RF lenses” in the future.
