The EOS R1 boasts sophisticated AF (autofocus) technologies. One of them is the newly developed cross-type AF, which improves subject focusing precision, providing more reliable and accurate AF in scenes that were previously challenging.

Canon’s Dual Pixel Intelligent AF system, which the EOS R1 adopts, uses the phase difference detection AF method, in which the image sensor's pixels made up of photodiodes are split into two, and the signals from each of the split pixels are used to focus the image.

On the EOS R1, this system is capable of cross-type AF, which performs horizontal line distance detection at a level on par with the previous method of vertical line distance detection, thanks to the proprietary pixel structure and array pattern. Cross-type AF’s capabilities are maximized by a newly developed algorithm capable of instantly calculating the best value based on the data obtained from horizontal and vertical line distance detection. In this way, the EOS R1 achieves more accurate AF than ever.

The Dual Pixel CMOS AF, a previous AF system, detects only vertical line distance. This led to focusing issues in scenes with few vertical elements such as birds with their wings spread out or subjects with vertical lines covering their faces such as players behind a volleyball net. The camera would be unable to acquire the subject, or lock on the vertical line patterns in front of the subject instead. Cross-type AF enables users to focus on any subject regardless of its shape, pattern, or any obstacles present.

Consistent AF is now possible in scenes with few vertical elements, and where subjects are behind obstacles with vertical lines such as nets.

Enhancements have also been made to tracking AF, which keeps the focus on a moving target subject. This was an issue when shooting dynamic team sports like soccer and basketball, where the AF would be thrown off by another player crossing the target subject during play.

Tracking AF involves two processes: acquisition (detecting the subject) and tracking (keeping the focus on the detected subject). On previous cameras, subject acquisition used deep learning (DL)-based processing, whereas tracking acquired subjects used a process that analyzed visible subject characteristics like shape, color, and brightness.

However, on the EOS R1, the use of DL-based processes was expanded to be applied to the tracking process as well. The tracking algorithms are pre-trained with a series of subject movements (i.e., feature data), which enables them to compare the feature data from the current and preceding moments during a shoot. As a result, the camera maintains precise focus even if the color, brightness, and shape of the target subject keep changing.

In the case of soccer and basketball games, players' movements often cross and the player being tracked temporarily disappears behind another player, so the aforementioned tracking method has to be suspended during that time. Even in such situations, the EOS R1 memorizes the facial features of the player during face recognition so that when the player appears again, it can identify the specific player among other players and resume tracking.

In a high-stakes situation at the goal line, where the leading player can change in an instance, the newly developed Action Priority AF feature proves its effectiveness. Unlike Tracking AF, which maintains the focus on a target player, Action Priority AF automatically switches the focus to a player performing a key action near the ball, making it easier to capture decisive moments such as shots, headers, and slides in soccer. Its new algorithm detects key actions by identifying and analyzing details such as ball position and athletes’ joint status, and can single out the player who will be the new focus of attention from among several other players.

The EOS R1 achieves a blazing fast 40 fps maximum continuous shooting speed that captures moments quicker than the naked human eye can see, as well as more accurate AF that stays tenaciously on a target. Such powerful features are possible with the support of a newly developed processing system.

There are two components that determine a digital camera’s performance: the image sensor, which receives and converts light into digital information, and the processor, which processes this information and creates the image file. These two components work together in real time with high precision to deliver high speed, high image quality, and precise AF.

The EOS R1’s newly developed image sensor has a faster digital data readout speed, producing a massive volume of data that must be processed by the processor at an unprecedented rate. Thus, the camera is equipped with a newly developed co-processor that works alongside the existing one to conduct split-second analysis, contributing to a leap in camera performance.

The EOS R1’s upscaling feature increases the resolution of images in-camera, catering to users who desire an increased pixel resolution for printing or cropping.

Most existing upscaling methods use color and brightness information from surrounding pixels to supplement pixels to enhance the sense of resolution needed for enlarging the image. However, this results in edge areas such as outlines becoming blurred or excessively emphasized. In comparison, the deep learning technology used in the EOS R1 faithfully predicts the pixel information in these edge areas, allowing the number of pixels to be doubled on both the horizontal and vertical axis without any change to the colors or brightness. The resulting image maintains a natural-looking resolution that looks more realistic.

The EOS R1 also features enhanced in-camera noise reduction capabilities. Subject details are often lost during existing noise reduction processes and there is strong demand for improvement among professional photographers. Deep learning technology makes it possible to identify the noise apart from the subject image data to remove noise without affecting the subject. In-camera noise processing can now minimize image noise without negatively impacting subject texture and detail.

The EOS R1 is a mirrorless camera developed to provide the ultimate performance and reliability. Canon’s top-of-the-line products, which have “1” in their product names, are equipped with advanced performance befitting flagship models. They inherit a legacy of Canon’s commitment to play a role in expanding the world of photography. Canon will continue to innovate and improve its technologies to expand imaging possibilities.

Recommended Articles

Read articles in ”Pioneering Technology”