Canon R5 Mark II Bird-in-Flight Settings That Actually Work

A swift doesn't telegraph its next move. One second it's cruising level over the water, the next it's banked forty degrees and dropped to snatch an insect off the surface, and whatever your autofocus point was locked to a moment ago is now aimed at empty sky. That's the test I run on any camera before I trust it for a full season of birding, and it's clearly the test Canon had in mind when they built the EOS R5 Mark II. I've had one in the bag since shortly after launch, mostly on estuary mornings chasing terns and a few weeks following ospreys and red kites on a river system that gets heavy raptor traffic in early autumn. What follows is what I've actually learned running it for flight work, not the version you get from a spec sheet.

What the Faster Sensor Readout Actually Buys You

The headline change from the original R5 isn't the megapixel count, it's still a 45-megapixel sensor, it's how fast that sensor gets read out. Canon markets the readout speed as roughly four times quicker than the first R5, and in practical terms that's the difference between an electronic shutter you can actually use for flight and one you avoid. On the original R5, shooting fast-moving subjects with the electronic shutter produced visible skew on wingtips, that classic jello-shutter smear where a bird's leading wing looks like it's been dragged sideways. On the Mark II, I've shot terns diving and gulls banking hard in a crosswind and the wing shape holds up. It's not mechanically impossible to see rolling shutter artifacts anymore, a hummingbird's wingbeat is still faster than any rolling readout can freeze cleanly, but for anything the size of a kestrel or larger moving at normal flight speeds, the distortion that used to rule out electronic shutter for birding is gone.

That matters because electronic shutter is silent, doesn't wear out a mechanical curtain, and on this body it unlocks a genuinely useful top speed. Whether you actually want that top speed for every situation is a separate question, and it's one I get wrong less often than I used to now that I've settled into a routine.

Autofocus Settings for Birds That Won't Fly in a Straight Line

The R5 Mark II's subject detection has an Animals category that further breaks down by species type, and birds get their own recognition model rather than being lumped in with dogs and deer. In good light against open sky it locks onto a bird's body and head reliably, and it will jump to the eye when the bird is large enough in the frame, a perched heron at 15 meters, say. At distance, where a tern is a few hundred pixels across, it tracks the whole bird shape instead, which is honestly what you want. Eye detection on a subject that's twelve pixels wide isn't doing anything useful.

Where I've had to actually tune things is the tracking parameters, not just leave it on Auto Case and hope. For fast, unpredictable fliers like swallows and swifts I run tracking sensitivity toward the responsive end, roughly +1, so the AF jumps to a bird cutting across the frame rather than hunting for what it was previously locked to. Acceleration and deceleration tracking I push toward the erratic setting too, because a swift changing speed mid-turn is exactly the case that setting exists for. The one I've learned to keep conservative is subject switching sensitivity. Set that too high near a mixed flock and the camera will happily jump off your target tern onto a gull that crosses six feet closer to the lens. I keep it low and just re-point the camera myself when I want to change subjects.

Eye Control AF is the feature everyone asks about, and for birding specifically I've found it more useful as a fast way to place my initial AF point on a perched bird than as something I lean on mid-flight. Calibration drifts if you shift your eye position relative to the finder, which happens constantly when you're tracking a bird overhead and craning your neck, and polarized sunglasses defeat it outright. I calibrate it, use it to snap onto a perched subject's eye in a fraction of a second, then let subject tracking take over once the bird lifts off. Treating it as a targeting shortcut rather than a tracking method has been the more reliable approach for me.

Frame Rate, Buffer, and the Card You Actually Need

The 30fps figure gets quoted everywhere and it's real, but it's not the setting I use for most of a session. Full-frame electronic shutter at 30fps fills the buffer fast and produces an enormous number of near-duplicate frames for every genuinely different wing position, which just means more culling later for marginal gain. Here's how the shooting modes actually break down in the field, based on a 325GB/s CFexpress Type B card and RAW capture:

Pre-continuous shooting is the one I undersold before I actually used it in earnest. It buffers a fraction of a second of frames continuously while you're half-pressing the shutter, so when a perched osprey finally commits to a dive, you've already got the wing-set from the instant before your brain registered it and your finger reacted. It won't save you from being pointed at the wrong bird, but for the specific problem of human reaction time on a launch, it's the single feature that's changed the most shots for me this season.

On card speed, don't cheap out. A slower CFexpress card or a UHS-II SD card in the second slot will let you hit 30fps for a second or two before the buffer stalls and the frame rate drops off a cliff mid-sequence, which is exactly the moment you don't want it happening. I shoot RAW to the CFexpress slot only during flight sessions and leave the SD slot for backup JPEGs, not simultaneous RAW.

Is the Crop Mode a Real Tool or a Checkbox Feature

The 1.6x crop mode reads out a smaller region of the sensor at full resolution readout speed, so you keep your frame rate and autofocus performance while gaining reach, at the cost of dropping from 45 megapixels to roughly 17. For a warbler or a sandpiper that's genuinely small in the frame even with a 500mm lens, that trade is worth it, you get more pixels on the actual bird even though the total file is smaller. What I don't do is leave it on by default. For anything where I've already got good subject size, an osprey at 15 meters with a 400mm lens, cropping just throws away resolution I'd rather keep for a print or a tight recompose later. It's a tool for a specific problem (subject too small in frame, can't get physically closer, don't have longer glass on hand) rather than a setting that should live on all the time.

Glass That Actually Pairs Well With This Body

The RF 100-500mm f4.5-7.1L is still my most-used lens for flight work and I don't think that's changing soon. At roughly 1.16kg it's light enough to track a bird handheld for an entire session without your arm giving out by mid-morning, and the R5 Mark II's autofocus stays confident even at f7.1 on the long end, which wasn't a given with older bodies. Add the 1.4x extender for 700mm at a usable f10 and it's still tracking birds in decent light, though I notice the hunting starts to creep in once cloud cover thickens.

The RF 200-800mm f6.3-9 is the one people ask me about most, because on paper 800mm of reach for around 2kg sounds too good. In practice it's exactly that, with the caveat that f9 at the long end eats a lot of light. On a bright morning it's fantastic for small, distant birds where the 100-500mm just doesn't have the pull. On an overcast afternoon you're pushing ISO hard to hold a shutter speed fast enough to freeze a wingbeat, and the R5 Mark II's sensor handles that reasonably well into the 6400-12800 range, but there's a real light floor below which this combination stops being fun to shoot with.

The RF 800mm f11 gets recommended a lot for budget reach and I'll say plainly it's not my pick for flight work. The fixed f11 aperture means you're already giving up a stop and a half compared to the zooms above before you've even started, and on a moving bird that's the difference between a shutter speed that freezes a wing and one that doesn't. It's a fine lens for a stationary perched shot in strong light. For anything airborne I'd rather carry the extra weight of the 200-800mm.

A Morning That Tells You More Than the Spec Sheet Does

The session that actually taught me most of this was an early September morning on a tidal estuary with a mixed flock of common terns working a bait ball maybe 40 meters offshore. I started on mechanical shutter at 12fps because that's my default, and for the first twenty minutes it was fine, terns diving at a fairly consistent angle and speed. Then a couple of Sandwich terns joined in, bigger, faster into the dive, and I kept clipping the moment of impact because 12fps wasn't giving me enough frames through that quarter-second window. Switched to electronic shutter at 30fps and immediately had the opposite problem, subject switching sensitivity was set too high from a previous session photographing a starling murmuration, and the AF kept jumping off my chosen bird onto whichever tern crossed nearest the lens. Dropped that setting down, left tracking sensitivity where it was, and the next forty minutes produced the best diving-tern sequences I've gotten with any body. The lesson wasn't really about the camera's ceiling, it was that the default settings from one kind of flight subject don't transfer cleanly to another, and I should have reset them before I started instead of mid-session.

Getting Through Four Thousand Frames Without Losing the Afternoon

That's the other side of a body that can do 30fps: a good two-hour session produces more frames than you can reasonably scrub through one by one in Lightroom before dinner. A single six-second dive sequence at full electronic shutter is 180 frames of a bird that, to a tired eye at 11pm, all looks roughly the same. This is the part of the workflow I've actually changed the most in the past year, and it's where imagic has ended up doing real work for me. It runs its sharpness and focus scoring locally on the machine, no upload queue, no waiting on a cloud service to churn through a card, which matters when I'm working off a laptop in a car park with no signal. It also clusters near-identical burst frames from a single flight pass together, so instead of scrubbing all 180 frames from that tern dive individually, I'm reviewing one cluster and picking the two or three where the wing position and eye sharpness actually landed.

The other piece that's saved me time is the apply_my_style preset, trained on my own past edits rather than a generic wildlife preset pack. Once I've picked the keepers from a session, applying my own established color and contrast treatment across the batch takes a few minutes instead of an evening of manually matching each image. None of this replaces actually knowing your gear, the culling software doesn't teach you tracking sensitivity settings, but if you want more detail on how the sharpness scoring and clustering actually work under the hood, I wrote about it separately in how AI photo culling works. And if buffer-filling burst sessions like this are a regular part of your week, it's worth reading through these workflow tips too, several of them changed how I structure a shoot day specifically because of how much data a body like this generates.

Frequently Asked Questions

Is 30fps electronic shutter actually necessary, or does 12fps mechanical cover most bird-in-flight work?

For the majority of flight photography, level cruising, gentle banking, birds approaching or departing a perch, 12fps mechanical is plenty and it saves you from wading through hundreds of near-identical frames later. I reach for 30fps specifically for short, high-speed events: a stoop, a dive, a fast wingbeat sequence where the peak moment might land between two 12fps frames. Treat it as a tool for specific bursts rather than a default shooting speed.

Does Eye Control AF actually help with fast, erratic fliers like swifts and swallows?

Not directly during the flight itself, in my experience. It's most useful for snapping your initial AF point onto a perched bird's eye almost instantly, before it takes off. Once a swift or swallow is airborne and changing direction constantly, subject tracking with tuned acceleration and switching sensitivity does more work than eye control can, and calibration drift from constantly repositioning your eye against the viewfinder while tracking overhead makes it less reliable mid-sequence anyway.

What CFexpress card speed do you actually need to sustain 30fps bursts?

You want a card rated well above the minimum, something in the 1700MB/s write range for CFexpress Type B, not a budget card that technically meets the spec floor. A slower card will hit its buffer limit within a second or two of sustained 30fps shooting and the frame rate drops noticeably mid-burst, which tends to happen exactly when the bird finally does something worth shooting.

Is the 1.6x crop mode a real substitute for a longer lens on small birds in flight?

It's a legitimate tool for a specific problem, not a substitute for actual focal length. You keep full frame rate and autofocus performance while gaining reach, but you're working with roughly 17 megapixels instead of 45, so there's less room to crop further afterward. I use it when a bird is genuinely too small in the frame with the glass I've brought, not as a default setting, since it throws away resolution you might want for anything larger and closer.