You also need to account for the desire for both dimensions in the 3:2 still images to be divisible by 8 for efficient JPEG compression.
True! I should also note this R5 camera will LIKELY have HEIF (High Efficiency Image File) format which can get down to 4x4 pixels for the DCT (Discrete Cosine Transform) blocks so that means the actual resolution can be a multiple of four. I did notice on internal RAW images that Canon TENDS to simply duplicate pixels on the edges to bring the resolution up to integer sizes. Again, I am pretty sure the TRUE native resolution of the sensor itself probably means it will have an extra 60 to 120 pixels on the vertical and/or horizontal axis used for internal black-levels and sensor noise calibration purposes.
---
I should ALSO NOTE Canon is NO STRANGER to high performance high pixel count sensors having made a 25,700 x 17,142 pixels 200+ mm super-sensor in 2009/2010, which would be a
440 Megapixel sensor + plus add extra calibration pixels to make it 448 megapixels -- So they KNOW how to make BIG SENSORS !!! And this sensor was from 2009/2010, so who knows what they've got in their labs NOW in the year 2020?!
Here is the original "TechNews Daily" magazine article for this sensor which is now being used in Satellite imaging:
MSNBC TECHNEWS DAILY: (2010)
Canon image sensor may redefine photography
updated 9/12/2010 3:33:41 PM ET
-Camera maker Canon recently announced what is by far the largest CMOS image sensor ever made, measuring 202mm x 205mm (8 inches x 8.1 inches). To put that in perspective, that's about 40 times the size (by sensor area) of Canon's next-largest CMOS sensor, the 35mm-sized (36mm x 24mm) sensor in the top-line Canon EOS-1Ds Mark III and EOS 5D Mark II digital SLRs. It's hundreds to thousands of times larger than the sensors typically used in point-and-shoot consumer still and video cameras.
The giant new sensor is also extremely sensitive and capable of extremely fast read-outs, which in turn allows it to shoot video images at high frame rates in very low light. Canon claims the ability to shoot video at 60 frames per second at light levels of 0.3 lux, which means it could easily shoot very good quality, extremely high resolution video in very low light.
But...why?
So, other than bragging rights, what could be the reason for building such an enormous image sensor? What would be the use?
It turns out there are endless uses for a sensor like this. Though Canon has not released the pixel count, it will obviously be capable extremely high image resolution. If it uses a pixel size of 9 microns x 9 microns (9 thousandths of a millimeter square — pure speculation at this point, but a typical pixel size for a professional digital camera) it would mean the new Canon sensor would be able to deliver approximately 488 megapixels per frame, far larger than any other previous single sensor.
That kind of resolution means, for instance, that a photo with this sensor encompassing all 102 stories of the Empire State building would be so detailed that one could make out the faces of every person looking out every window. (Each pixel would cover an area of 16mm x 16mm, or 2/3" x 2/3".) And because the new Canon sensor also boasts very high (0.3 lux) sensitivity and very high (60 frame per second) frame rates, it could show those faces in real time on high-speed video. By moonlight.
Nothing even near that kind of resolution and capability has ever been possible with any previous image sensor.
Don't wait for the pocket camera! According to Canon, the limitation on the size of this sensor is that CMOS silicon wafers (the thin silicon-based disks of which computer chips are made) are currently not large enough to make CMOS sensors any larger. And because that greatly limits the production of sensors like this, the cost of these sensors will always be very, very high, quite possibly in the hundreds of thousands of dollars each. Adding to the cost of a camera for this sensor will be the optics required to make use of it. The very large, extremely high quality optical elements required to use the full resolution of the
sensor could easily cost more and take longer to make than the sensor and support electronics. So cameras using this sensor are not likely ever to be produced in large numbers.
So who will use them?
Use of these sensors will likely be limited to scientific, military, or other specialized applications where such costs can be justified. Astronomers will likely be among the first to use them as sensors for sky surveys where the requirement is for a sensor that can cover a relatively large area of sky while still maintaining high resolution. Aerial mapping, forest and geological surveys, sky surveillance, ground-based satellite tracking, and other such very specialized uses will likely be where these sensors are used, though one can't help wondering what kind of cityscapes, landscapes, and high-resolution movies could be
made with something like this.
But even if they're never available in large numbers, this breakthrough image sensor is likely to redefine what's possible in many forms of imaging. And because the nature of digital electronics is that prices are driven ever downward, it will also likely lead to much improved sensors and technology, potentially all the way down to popular commercial and even personal cameras, making it a very significant breakthrough indeed.
----
