The EOS R1 becomes the official flagship in Canon’s mirrorless series. A specialist camera aimed primarily at pro sports photographers, it features a 24.2 Megapixel stacked sensor, 40fps electronic bursts with pre-capture options, an impressive viewfinder with eye-control, 6k video, and their most advanced autofocus to date, all packed into a tough body with a built-in portrait grip.

Launched in July 2024, it comes four years after the 1Dx Mark III, the previous model in the prestigious 1 series. Canon brought me to the European press event in Munich Germany, where I got to try out the R1 and R5 Mark II launched alongside it – check out my EOS R5 Mark II review so far if you’re interested.

In the video below I’ll show you what’s new, what features are shared between the two bodies, along with some results from these pre-production cameras. I’ll follow-up with my full in-depth reviews once final production models are available, and will link to them here when they’re ready. So while this video is a first-looks report, I’ve still got loads to show you. PS – if you prefer to read the written highlights, keep scrolling!

Looking back at recent 1 series launches, Canon’s kept to a roughly four year cycle, in turn coinciding with the Summer Olympics. As such it was no surprise for the R1 to be launched in mid-24, but given the three year old EOS R3 was already so capable, I wondered where Canon could go for its flagship model?

Most obviously many of us assumed it would debut Canon’s first sensor with a global shutter, especially since they’ve been teasing developments and arch rival Sony has already played its hand with the A9 III. But the R1 has arrived with a stacked sensor instead, admittedly faster than the R3’s, but still lacking global readout.

Canon explained they can achieve better image quality right now with their stacked sensor, coupled with a new low pass filter to maximise detail, while the improved readout is sufficiently fast to minimise skewing from rolling shutter. 

As for the resolution, Canon’s been pitching the 1 series at around 18-24 Megapixels for many generations now, so clearly believes this is the right amount for pro sports. If you want more, go for the R5 II launched alongside it, which again shares many of the same features.

Of course any company will claim their current solution delivers the optimal results, but you do wonder if the R1 was originally slated to include global, and that it simply wasn’t ready for the Paris Olympics, or even if it turned out to be inappropriate technology for their flagship sports camera. 

Interestingly there was a global sensor upgrade option for the C700 cinema camera, but that’s designed for video alone and lacks autofocus, so isn’t suitable for a camera aimed at high-speed sports capture.

Analysing results from the R1 will of course show where it stands against the competition, but the lack of a new sensor type does leave the R1 without an obvious headline feature to shout about. 

Indeed at first glance it appears to share so much with the R3 that Canon’s presentation included a slide listing 100 differences and improvements. Some are certainly bigger than others, but the R1 is more about the sum of these parts rather than a handful of hero upgrades.

Right, let’s get into the design which again at first glance looks a lot like the existing EOS R3, albeit with a different textured finish. 

1Dx Mark III: 158x168x83mm, 1250g // EOS R1: 158x150x83mm, 1115g // EOS R3: 150x143x87mm, 1015g

The R1 is a tad larger, 8mm taller and 7mm wider than the R3, not to mention 100g heavier, but unless you have them side-by-side, you’re unlikely to notice much difference in your hands.

While the R1 does become closer to the 1Dx Mark III in overall heft, the DSLR remains noticeably taller and 135g heavier, so there are some mild savings with the mirrorless version.

Like the R3 and 1Dx Mark III, the R1’s build quality is fantastic, feeling almost bullet-proof and able to take on whatever you throw at it. I know Sony prefers to let you decide whether to fit an optional grip, but having it built-in just gives these bodies an undeniable air of confidence, solidity and robustness – and I love how they look and feel.

Suffice it to say the R1 is fully weather-sealed and while there don’t appear to be any obvious vents that I could see, I didn’t experience any overheating in my tests so far. Like the R3, the sheer heft of the camera and the chance to better separate the components, battery and slots improves heat dissipation.

From the top, the R1 looks almost identical to the R3, starting with two buttons to adjust the drive and AF modes, bracketing, flash compensation and metering.

In the middle is one of Canon’s multi-function shoes first seen on the R3, but now sporting a new cover with a clip to hold it place. The fastest flash sync is 1/320 with the EFC shutter or 1/400 with the fully electronic – not bad, although way below the A9 III.

Meanwhile the upper right side is dominated by a square information screen, alongside a button that reverses the text and background colour. As before there’s a mode button inside the thumb dial, along with an exposure compensation button, but new to the R1 is one for white balance too.

Meanwhile towards the front of the grip are a finger dial, the m-fn button for adjusting AF area and the soft touch shutter release. These, along with the joystick, smart controller and some corner buttons, are all duplicated when holding the camera by the portrait grip.

Speaking of which, look below the body and you’ll see Canon’s also included a strap lug that lets you carry the camera in the portrait orientation – a feature from the 1Dx III that was sadly lacking on the R3.

Oh and as teased, the R1 and R5 Mark II become Canon’s first bodies that can use the aperture ring on hybrid Z lenses for photos as well as video. I believe this will also come to future bodies, but I’m unsure about older ones.

Turning to the rear, the control layout again looks much the same as the R3, including a collar to switch between movies and stills with a record button in the middle, the choice of a joystick or smart controller, the traditional large flat thumbwheel and a clicky power switch with a lock position.

New to the R1 over the R3 are a duplicate info button for portrait shooting, and an extra button under the screen, allowing Rate in the top left to be reassigned as a spare function button. 

Sadly unlike the 1Dx Mark III, none of these buttons are backlit.

Like the 1Dx Mark III, some of the rear buttons can also be backlit for easier visibility in low light.

If you’re not familiar with the R3 or 1Dx III, you’ll be wondering what exactly is the Smart Controller alongside the joystick. This works as a standard AF-On button, but also provides an alternative way to move the AF area or push around an enlarged image in playback. It employs optical technology, kind of like an upside-down computer mouse, and I quite enjoy how it feels under your thumb.

But the R1 extends its functionality by supporting a half-press action, which you can configure separately to a full push from the menus if desired.

Dominating the rear of the camera though is the substantial housing for the viewfinder which includes an improved version of Canon’s eye-control technology first seen on the EOS R3.

But first the EVF itself which becomes Canon’s first to employ a 9.44 million dot panel coupled with a generous 0.9x magnification. This is not only larger and more detailed than the R3, but brighter too, making it by far Canon’s best quality electronic viewfinder to date. There’s also an option for blackout-free shooting, albeit I believe with a tiny delay.

The large housing not only provides a comfortable shade and anti-fog features, but also accommodates the infra-red LEDs and sensors used for eye-control, which can track the position of your actual eye and know where you’re looking.

Like the R3, the idea is not to position an actual AF area by eye-control alone, but to select between subjects that have already been recognised by the AF system by simply glancing at them. 

This worked really well for me on the R3, and now on the R1 and R5 Mark II, Canon has improved the accuracy. You can also refine the system by repeating the calibration process, and can store results for multiple people.

I found it surprisingly effective at quickly selecting a target for the AF system to subsequently track, which makes it ideal whether you’re photographing sports teams, groups of birds or animals, or people at an event or even in a street scenario. 

It may not work in video, but remains a valuable and unique feature in the Canon system. And as I’ll show you in a moment, further AF advances make it even easier to drill down to the desired target.

Moving onto the screen, the R1 employs a 3.2in panel with 2.1 million dots that’s side-hinged, allowing it to flip out and twist to any angle including forward to face you or back on itself for protection. 

Since the 1Dx Mark III had a fixed screen, I wondered if Canon would follow-suit with the R1, but they clearly believe the mechanism used on the R3 is sufficiently robust, although strangely the panel itself is less detailed than the 4.15 million dots of the R3.

Like the R3, the R1 is powered by the same LP-E19 pack, rated for 1330 shots with the screen or 700 with the viewfinder under CIPA conditions, more than quoted for the R3, although in practice you’ll probably achieve a lot more on either camera especially if you’re firing-off electronic bursts.

Behind a door on the grip side, you’ll find twin CF Express Type B card slots, the same as the 1Dx Mark III, but a key difference over the R3 and R5 II which have one CF Express and one SD slot.

Behind a series of rubber flaps on the left side are 3.5mm microphone and headphone jacks, flash sync, USB-C and full-size HDMI ports, and an ethernet port that now supports faster 2.5Gbit speeds.

It’s great that Canon’s finally got the message about unwanted Micro HDMI ports, and you can use the R1’s to also output ProRes RAW video. Meanwhile the USB-C port supports v3.2 and can work as a webcam, although sadly there’s no option to connect a USB SSD drive.

I am pleased to report the R1 inherits the built-in GPS receiver of the R3, allowing you to tag images with location data and local time without any additional accessories. You can see the embedded data here.

Canon’s also upgraded the Wifi to support 6E, allowing it to connect to less congested 6GHz networks.

Moving on, as you already know, the EOS R1 employs a back-illuminated stacked CMOS sensor with 24.2 Megapixels, but Canon stressed this was different from that in the EOS R3.

For starters it’s faster than the R3, with Canon claiming the rolling shutter on the R1’s electronic sensor is in fact now faster than the mechanical shutter on the 1Dx Mark III.

Here’s a quick panning test with the electronic shutter at 40fps where there’s no visible skewing on the vertical lines, or at least none that I’m worried about just yet. I’ll be doing more tests in my final review as well as comparing it against rivals.

Secondly, the R1’s sensor may share the same 24 Megapixel resolution as the R3, but is coupled with a new low pass filter that claims to reduce colour and brightness moire while improving the resolved detail.

The R1 can also boost the resolution, but not with a pixel shift mode as you might expect. In fact while the camera does of course have IBIS which can move the sensor with up to 8.5 stops of compensation, there’s no pixel shift option on the R1 or R5 II.

Instead for both cameras, Canon’s opted to implement AI Upscaling in playback on any JPEG or HEIF file you’ve captured. This doubles the pixel-count in each axis, generating a file with 96 Megapixels in-camera. It took about 5 seconds to process a JPEG in my tests.

Canon’s opted to exclude in-camera upscaling of RAW files as it would take too long, but I believe this will be offered in an upcoming plugin for Lightroom.

So let’s have a look at some examples, which I took with a pre-production camera, first zooming-in on the original 24 Megapixel file before comparing it to the upscaled version on the right.

As you can see, the upscaling is delivering what looks like an intelligently sharpened image with the appearance of more detail in some areas. I believe the process may be recognising certain elements and enhancing them differently to others, and when looking closely at my samples, with varying effectiveness.

Ultimately your mileage will vary depending on the subject but importantly none of this is baked-into a pixel-shift file, nor subject to the motion issues that plagues that particular process. You take your photos as normal, then choose whether to generate an enhanced version in playback later.

The benefit of Canon upscaling in-camera, versus a third-party on a computer later, is having knowledge of the entire capture pipeline from the lens and camera type to the image processing that’s already been applied and in theory can be reversed. You can also select multiple images in playback to process in bulk. 

Going forward, I’m interested to see how enhancing RAW files might avoid having to deal with pre-compressed JPEG and HEIF images, although Canon assured me the process is aware of how an image is compressed and can effectively reverse it.

Canon’s also been talking about noise reduction with neural networks for some time, and this too has now made it onto the R1’s playback menus, but this time as part of RAW post processing.

So when you select a RAW file for processing during playback, you’ll not only see the usual options, but under the NR section, a new setting to enable neural network noise reduction. As before, you select the options you desire, before saving the result as a new JPEG or HEIF file, which of course can subsequently be AI up-scaled in-camera too if you like.

To show you what it can do, here’s a RAW+JPEG I took with the R1 at 12800 ISO, so let’s zoom into a portion for a closer look. I’ll keep the original in-camera JPEG on the left, before comparing it to the RAW version on the right processed with Neural Network noise reduction set to the default middle value. 

At this sensitivity there already wasn’t much noise on the original to start with, but the processed version on the right is visibly cleaner with better-defined edges. In my tests with the R5 II, the benefit became more apparent at higher sensitivities.

There’s also a new Blur Detection feature which can tag images in playback to help you more quickly identify those which are sharp.

Moving onto autofocus, the R1 enjoys a number of upgrades, aided by a new DIGIC Accelerator processor, and many of the new features are shared with the R5 II. I’ll detail them in a moment, but first talk about one that’s exclusive to the R1: it sports Canon’s first imaging sensor with cross-type autofocus, improving accuracy on vertical detail. 

In Canon’s Dual Pixel system, each colour filter in the traditional two-by-two Bayer array covers a pair of photo diodes, allowing them to perform both focus and imaging duties. Now under one of those four colour filters, actually the second of the two green filters, the pair of diodes are twisted by 90 degrees.

As for what’s shared with the R5 II, subject recognition now includes horses, aeroplanes and trains, while human recognition has received major enhancements. First the system can now better identify poses and upper bodies, stay locked-onto a subject as others pass in front of it, as well as recognising actual sporting moves. 

The new Action Priority mode has knowledge of three sports, soccer, basketball and volleyball, and attempts to anticipate the movement of the ball to change the target AF area. 

In short, if it recognises one player passing to another, it should automatically switch to the recipient. Or if it detects, say, a spike in volleyball or an approaching dunk in basketball, it’ll again switch to that player. 

You may be wondering why these three sports in particular? Well right now, the technology is looking for the movement of one, large circular ball, as it moves between players, as opposed to smaller or non-circular balls that may be harder to see or held by a player. Other sports may be added in the future though.

You can further refine the priority of specific people by registering their faces with the camera. You can pre-load a photo from any source, or take one with the camera itself, and amazingly it only needs one shot face-on to work its magic. I believe you can store ten different people in a bank and you can store up to ten banks on a card, although only one is active in the camera at a time.

Here I’ve registered myself with a single photo, after which you can see an additional icon alongside me in live view, and see how it stays recognised even as I turn to profile. Note how I’ve also restricted it to a central zoned area.

This can not only be useful if you’re concentrating on key players in a match, but equally beneficial at events or weddings where you can pre-load images of the most important people.

I tried out the R1 for shooting basketball at the Canon event and found that Action Priority did a good job at anticipating the action most of the time, when given the entire frame to analyse. I also found the R1’s AF felt snappier and more engaged than the R5 II, presumably aided by those cross-type sensors.

As with all AF systems, you’ll improve your success if you can drill-down, traditionally with a more focused area, but here by exploiting registered people, or using the eye-control to over-ride and quickly select the desired target.

I’ve only had a brief time with the R1 so far, but the AF system is definitely a highlight, as indeed it should be on a flagship sports camera.

As for burst shooting, the R1’s mechanical shutter offers the same 12fps at up to 1/8000 as the R3, but the electronic shutter has accelerated to a top speed of 40fps at up to 1/64000. 

In a small but very useful upgrade, you can also select slower frame rates at more intervals, including 30, 20, 15, 12, 10, 7.5, 5, 3, 2 or 1fps. 

I couldn’t find the limited high-speed burst mode of the R3, but more importantly the R1 does now have a proper pre-burst option when shooting electronically.

It works as you’d hope, buffering up to 20 shots as you half press the shutter, which works out at half a second’s worth when set to the fastest 40fps, or a whole second at 20fps. 

Half a second doesn’t sound like much, but it’s actually plenty if you fully push the shutter down as soon as the key event takes place, like a ball being kicked, or a bird taking flight. These sequences were all taken with pre-burst.

Unlike the R5 II which became quite warm during sustained bursts with pre-capture though, the R1’s temperature seemed to remain pretty constant throughout my tests – again the larger body provides room to separate components and better manage the heat.

In some of my bursts, you may notice my pre-production R1 exhibiting some banding when shooting directly into some very bright artificial lights. I also noticed this with the R5 II under the same conditions. I’m not yet sure if the electronic shutter was to blame, and I should note it didn’t affect my video clips. Much more testing is required under a variety of conditions.

Moving onto video, the R1 builds upon its predecessor’s capabilities with a number of options that better match the cinema camera series. 

These include renaming MP4 to XF-AVC S or XF-HEVC S, storing them in a new folder, and adding C-Log 2 as well as 3 in a new custom picture settings menu.

The R1 may miss out on the waveform monitor of the R5 II, but does support for four channel audio and a tally lamp by the logo.

You can film 1080 up to 240p or 4k up to 120p in either traditional 16:9 UHD or wider DCi formats, but there’s still no open gate option. A 4k Fine mode oversamples from 6k up to 60p, while 6k DCi itself is available up to 60p in Canon’s CRM RAW format, or output over HDMI as ProRes RAW. 

Pre-capture for video is available with a three or five second buffer, and there’s also a new mode that lets you shoot stills as you record video, albeit limited to 1080 video with 17 Megapixel stills in the 16:9 shape.

Graphite sheets for the sensor and card slot work with an aluminium alloy plate to dissipate heat, and Canon reckons you should get at least two hours worth of recording in most modes up to 60p. Rest assured I will be doing plenty of overheating tests in my final review, but in the meantime, the R3 was only limited by memory or power in my long recording tests.

And that’s all I can say for now until I spend more time with a final production model. Initially I was disappointed there was no killer headline feature on the R1, but in my time with it so far, I found it an extremely capable camera none-the-less.

I think its biggest problem is the R3 was already so good it arguably deserved the 1 badge three years ago. That in turn would have made this new camera spec more comfortable being pitched as a Mark II version rather than becoming the first mirrorless 1 series.

Then there’s the global shutter question, but what you need to really ask is whether you actually need it? Based on my A9 III tests, they certainly have unique benefits over conventional sensors, not just in eliminating skewing, but also avoiding issues with banding while allowing very fast flash syncs. On the downside, they do seem to come at the cost of reduced dynamic range, increased noise, and the high price of the sensor itself.

So while it’s interesting to speculate whether Canon originally intended for the R1 to have a global sensor and perhaps didn’t achieve the quality it desired in time, the bottom line is to ask whether the R1 will give you the results you desire regardless, or if the alternative tech in the A9 III, or indeed a Nikon Z-9 or an older R3 will be more appropriate. 

Check back later for my final review, but if you’re already sold on the R1 – or it’s cemented an alternative decision, please do consider ordering through my links below. I also have a separate review all about the R5 II if that’s more your cup of tea.