You leave before the light does. Eight degrees, a fine drizzle that does not look like rain, the road surface dark and reflective. Twenty-two kilometres later you are home with wet gloves and a card full of footage that looks nothing like the morning you just rode.
The footage is grey. Not dark — grey, flat, as if someone turned the contrast down. Some of it shakes in a way it does not shake in July. The last six clips are missing entirely, because the camera shut itself off at the top of the second climb with the battery reading one bar.
None of that is a skill problem and none of it is a bad camera. Three physical quantities went out of range on the same morning, and they went out of range together. This is the article about that: not how to shoot beautifully in bad weather, but why winter takes your footage apart, and which parts you can actually do something about.
The light arrives after you do. Most winter footage problems are decided before the first pedal stroke.
Rain, darkness and cold are usually sold as three separate problems with three separate accessories. They are not separate. They are one chain, and the chain runs in one direction.
Cold slows the battery. A pack that gives you eighty minutes at twenty degrees gives you thirty-five at four. That is not a fault; it is chemistry, and it happens before anything else does.
Less battery means fewer minutes. So you shoot less, or you drop from 4K60 to 1080p, or you turn off the thing that was holding the image steady. Every one of those decisions is made while standing next to a bike in the cold, and every one of them costs quality later.
Darkness then removes the margin. The settings you fell back to in order to save power need light. Without light they produce noise, then motion blur, then nothing usable. This is where the causal chain becomes visible in the edit.
And water sits on top of all of it, quietly, doing the one thing post-production cannot fix: interrupting the light path between the world and the sensor.
So the useful mental model is not "bad weather gear". It is three curves — light, charge, water — that all bend downward at the same hour of the day in the same four months of the year. Understanding where they cross tells you what to fix and, more importantly, what not to bother fixing. The same three things fail on a six-hour ride, just on a slower clock — that version is in the long-ride piece.
Everyone worries about the camera going under. Almost nobody worries about the millimetre of water sitting on the front element, which is the thing that actually ruins the shot, and it does it on every single clip rather than once.
A clamp is a gap: water enters, grit stays, and the joint starts to drag instead of locking.
A droplet is a lens. Water has a refractive index around 1.33. A bead of it on flat glass bends the light passing through it, so part of your frame is now looking through a small, badly shaped, uncoated lens. The result is not a blurry picture — it is a picture where one region has lost contrast, developed a bright halo, or smeared the highlights of a wet road into a white band. No amount of sharpening in the edit puts that detail back, because it was never recorded.
IPX4 is not what you think it is. That rating means splashing water from any direction causes no harm. It says nothing about immersion, nothing about pressure, and nothing about the joint between the front glass and the housing, which is exactly where water collects when you ride at thirty kilometres an hour into drizzle. The camera will survive the ride. The image will not survive the droplet.
Where water actually hurts the hardware is the seams. A rotating head has a gap. A quick-release collar has a gap. A clamp screw has a thread. Water gets into all three, and what it brings with it is grit — road spray is a suspension of fine mineral particles, and a wet joint is a lap. Once that paste is inside a ball head or a ratchet, the mechanism does not fail suddenly; it starts to drag, then to seize at one angle, then to stop seating fully, which shows up as a camera that has quietly rotated three degrees since you set it.
Salt is the accelerant. If your roads get gritted, you are riding through calcium chloride, and it is hygroscopic — it pulls moisture out of the air and holds it against metal long after the road looks dry. Stainless steel fasteners threaded into aluminium under those conditions produce galvanic corrosion, and a screw that came out by hand in November can need a tool, and then a drill, by February.
Then it freezes. Water in a thread or a collar expands when it turns to ice, and a gap that was designed to be a few hundredths of a millimetre becomes a gap that is being pushed apart from the inside. Do that twenty nights in a row and the part that was precision is now loose.
The single most common winter complaint is "the battery died in twenty minutes". The battery did not die. It became a worse battery, temporarily, and understanding the difference saves you both money and footage.
What cold does to lithium cells. Below about five degrees, the internal resistance of a lithium-ion cell climbs sharply, and at minus ten it can be two to three times its warm value. Higher resistance means more of the stored energy is lost as heat inside the cell and less reaches the terminals, so the voltage sags under load. The camera sees a sagging voltage and concludes the pack is empty. Usable capacity at minus ten is commonly half of what the label says; at minus twenty, a third.
Soft pads are the whole grip. Hard ones still look like pads and no longer act like them.
Which means the shutdown is a lie you can undo. Warm that same pack to room temperature and it will read forty percent again. The energy was there the whole time. It was just not deliverable at that temperature.
So insulate the battery, not the camera. This is the counter-intuitive part, and it is where most people spend money in the wrong place. An action camera generates its own heat while recording — a small body with no fan gets warm within minutes. Wrapping it in foam makes it hotter, which shortens its life and does nothing for the pack. The pack, on the other hand, has no heat source at all, and it is the component that fails. Keep spares in an inside jersey pocket against your back. Rotate them. A pack at thirty degrees behaves like the specification; the same pack on a cold down tube behaves like a rumour.
Cold also changes materials, and this is where mounts get blamed for something they did not do. Silicone and rubber pads harden below freezing; the coefficient of friction drops and the pad stops conforming to the tube. A clamp that gripped perfectly in September will rotate on a frosty morning at the same torque, and the rider concludes the clamp is rubbish. It is not. The interface changed. Similarly, the polycarbonate and ABS parts in a plastic mount lose thirty to fifty percent of their impact strength in the cold, which is why the failures you see in winter are brittle snaps rather than slow deformations. The clamp-versus-stem-cap trade-off in the buying guide is worth re-reading with that in mind.
And the screen stops cooperating. LCD response slows, touch becomes unreliable, and you are wearing gloves. Any part of the workflow that requires two hands and a menu is a part you will skip, which means the settings you intended to change never get changed.
The instinct in low light is to raise ISO and keep shooting. ISO does not add light. It multiplies a smaller signal, and it charges interest in the shadows — which in cycling footage is precisely where the important information lives.
Low light is where a stabilised image stops being sharp and starts being smeared.
What you actually lose. Every doubling of ISO costs roughly one stop of dynamic range. In a dark winter scene the histogram is already squeezed into the bottom third; pushing ISO moves the whole thing up but stretches the noise floor with it. The road surface, the tree line, the inside of a jersey — those are your shadow regions, and they are the first to turn into coloured mush. Highlights, meanwhile, are still highlights: a wet road under a car headlight will clip no matter what you do.
The shutter is the real casualty. Low light forces a slower shutter, and a slower shutter converts every small vibration into motion blur. This is the counter-intuitive bit about stabilisation: electronic stabilisation works by cropping and shifting the frame, and it needs a reasonably clean image to track. In good light it can rescue hand-shake. At dusk, with a slow shutter and heavy noise, it is trying to stabilise an image that is already smeared, and it makes the smearing worse by resampling it. The gyro-based systems do better because they act on the readout, not the pixels, but none of them can invent detail that was blurred at capture.
So drop the frame rate, not the resolution. Everyone shoots 4K60 in summer because it is free in daylight and it gives you slow motion. In winter it is not free: doubling the frame rate halves the light per frame at a fixed shutter angle, which pushes you towards either a wider aperture you do not have or a higher ISO you cannot afford. Thirty frames per second at a wider shutter is the better trade for a dark ride, and it will look cleaner on a phone screen than 4K60 that has been denoised into plastic.
The one nobody warns you about: flicker. LED bike lights and street lighting are usually dimmed by pulse-width modulation, switching on and off hundreds or thousands of times a second. A rolling shutter reads the sensor line by line over the course of a frame, so a light that is pulsing faster than the readout produces horizontal bands of alternating brightness. It shows up as stripes crawling up your footage, and it is worst exactly when you want the shot — a dark road, a bright light, a slow shutter. The fixes are unglamorous: shoot a test clip at the light you plan to use, change shutter speed until the bands disappear or become static, and avoid pointing the camera directly into a modulating light source.
Individually, each of these is manageable. Together they produce the only genuinely hopeless combination in winter shooting, and it is worth knowing why so you stop fighting it.
At four degrees with rain after dark, you have: a battery delivering roughly half its rated time, a lens that collects droplets faster than you can clear them because your gloves are wet, and a scene whose shadows are already at the noise floor. Every mitigation for one problem makes another worse. Wiping the lens adds a smear. Raising ISO to see the road adds noise to a scene that has no margin. Shooting longer clips to save battery means more footage through a wet lens.
The practical answer is to change what you are shooting rather than how. Reduce to short, deliberate clips at a fixed position. Pick one place with a light source and let the ride come to it. Use a static frame and let motion happen inside it — water crossing the frame, wheels passing, a light sweeping through. A twenty-second locked-off shot in the rain at night is worth more than two minutes of handheld searching, and it costs a tenth of the battery.
The other answer is to accept that this is a stills-and-short-clips evening, and that the ride itself was the point. Cameras are tools for recording; they are not a requirement for having gone.
When the conditions stack up, stop chasing the ride and let it cross a fixed frame.
This is the part nobody puts in a specification table, because it is not a feature. It is a list of ways a mount fails at four degrees that it does not fail at twenty.
Differential contraction. Aluminium contracts about a third more than steel per degree. A stainless screw in an aluminium body, tightened warm, loses preload as the assembly cools — not dramatically, but enough that a joint which tested tight indoors feels different outdoors. That is why the first check of the morning should be done in the cold, not in the hallway.
Steel screw, aluminium body, salt on the road: the fastener becomes the failure point.
Rubber that stops being rubber. The pad is the whole grip. A hardened pad has less contact area, and a clamp cannot compensate with more torque because the tube underneath has a limit — carbon bars in particular are typically rated around five to six newton-metres, which is less than most people's instinct. In the cold, the correct move is a softer, wider pad and a clean tube, not a longer spanner.
Mechanisms with springs. Quick-release catches, spring-loaded collars and detent pins all rely on a spring doing its job against cold, stiff grease. A catch that does not fully seat looks closed, feels closed, and releases under vibration. If a locking mechanism needs two hands and a warm day to work properly, it is the wrong mechanism for winter.
Fewer joints is better than better joints. Every rotating interface is a place for water to enter and a place for a degree of drift to accumulate. A mount with one well-sealed adjustment will hold its aim through a winter; a mount with four will be a different angle every week, and you will spend the season re-aiming instead of riding. It is the same mechanism behind most bad bike footage, just slower.
One sealed adjustment holds its aim all winter. Four of them hold nothing.
Metal where it matters, and the right metal. The contact surfaces that carry load should be metal, because plastic creeps under constant load and cracks under impact when cold. But dissimilar metals in a salt environment need to be chosen deliberately, or the fastener becomes the failure point instead of the mount.
Glove-sized controls. Whatever the mechanism is, it has to be operable with a wet glove in the dark. That is a design constraint with a real cost, and it is usually the first thing traded away for a cleaner look.
This is the whole preparation routine. It takes ten minutes, it needs no tools, and it happens standing next to the bike in the cold — which is the point, because the cold is the test condition.
Ten minutes at the door, in the cold, with gloves on. That is the whole checklist.
1. Three batteries, two of them on your body. One in the camera, two in an inside pocket. Not in a saddle bag, not in a frame bag, not in a top tube bag. Against your back.
2. Two cloths, not one. One for the lens, one dry one for the housing. A wet cloth smears; a smeared lens is worse than a wet one because you cannot wipe it off.
3. Press the pads with a thumbnail. If they do not deform, they are hard, and the clamp will slip no matter how tight you make it.
4. Check the threads from yesterday. If the bike was ridden wet, the collar has water in it. Dry it, and put a thin film of grease on anything that rotates.
5. Let the camera acclimatise, or seal it. Taking a warm camera from a warm house into cold, damp air puts condensation on the outside and, over time, inside. Give it five minutes in a hallway or a bag before you expose it.
6. Set exposure before you leave. Winter scenes are high-contrast: dark road, bright sky, wet specular highlights. A touch of positive exposure compensation protects the shadows where your detail is; you can pull the sky back later.
7. Screen brightness up, touch off. You will not be operating a touchscreen in gloves. Set the mode you want and lock it.
8. Format the card. Every ride. It is faster than deleting and it prevents the write error you will otherwise discover at the best moment of the morning.
9. Fit the tether. Cold fingers drop things, and a dropped camera on wet tarmac is a finished camera.
10. Aim, then stop aiming. Set the angle in the cold with gloves on, then leave it alone. Every subsequent adjustment is done with wet hands and costs you a clip.
Wipe at every stop, not at every problem. You stop anyway — a light, a layer, a junction. Use those twenty seconds. A lens that is cleaned before it is visibly dirty never reaches the state where you are trying to polish a smear with a wet sleeve.
Wipe at the stops you were making anyway. Twenty seconds, one dry cloth.
Rotate the battery before it dies, not after. Swap at the halfway point of the expected life, while the pack still has charge. A cold pack that has been drained to shutdown needs to warm up before it will accept a useful charge again, and you will not be doing that at the roadside.
Do not breathe on the glass. It is the standard trick and it is the wrong one in winter: warm breath condenses instantly on cold glass, and the moisture you add is the moisture that later freezes or fogs from the inside. Use the dry cloth.
The most damaging thing people do to a wet camera happens in the first ten minutes at home, and it is done with good intentions.
Do not use a hairdryer or a radiator. Heat drives moisture inward, past seals, towards the sensor and the lens groups where it cannot escape. A camera that would have dried overnight becomes a camera with fungus in three weeks.
Do not power it on to "check". If there is condensation inside, powering up is the moment it can short. Take out the battery and the card, leave the ports open, and let it sit at room temperature for a few hours.
Rinse the salt, dry the threads, re-grease. Every time, not once a season.
Fresh water, not salt water. If the roads were gritted, the residue on the mount is salt, and salt keeps pulling moisture out of the air. Rinse the hardware in lukewarm fresh water, dry it immediately, and re-grease every thread and every rotating interface.
Inspect the pads and the plastics. A pad that has gone shiny and hard will not recover. A plastic catch that has been flexed cold a few dozen times is a catch with a crack you cannot see yet. These are cheap parts; replace them before they cost you a camera.
Then leave it alone. The right storage for winter hardware is dry, room temperature and slightly over-greased. Not warm, not sealed in a bag with the water it brought home.
Can I shoot in the rain with a normal action camera? Usually yes, for splash and drizzle. The body is typically rated for it; the lens coating and the seams are the weak points. Assume the camera survives and the image does not, and plan your wipes accordingly.
Do I need a waterproof housing? Only for immersion or sustained heavy rain. For drizzle, a housing adds a second piece of glass in front of the lens, which is a second surface to get wet and a second surface to fog. Most winter footage problems are made worse by a housing, not solved by it.
My lens fogs from the inside. Why? Moisture entered when the camera was warm and damp, then condensed when it cooled. The fix is preventive: acclimatise before you go out, dry before you come in, and never store it sealed while wet.
My battery reads empty but works again indoors. Is it broken? No. Cold raised the internal resistance and the voltage sagged under load. Warm it, and it will deliver most of its capacity. Two packs rotated against your body is a better investment than a bigger pack.
How do I stop the stripes from bike lights at night? Change shutter speed until the bands disappear or stop crawling, and avoid aiming directly into a modulating LED. Test one clip before you commit to the shot — it takes ten seconds and saves the whole evening.
Is a clamp safe on a carbon bar in the cold? At the manufacturer's torque, with a clean tube and a soft pad, yes. The danger is compensating for a hardened pad by tightening harder. If the clamp needs more force to hold, the pad is the problem, not the torque.
Winter does not ask you to be a better filmmaker. It asks you to be organised before you leave and honest about what the conditions will allow. The riders who come back with usable footage in January are not shooting more — they are shooting less, in better places, with dry glass and a warm battery.