Choosing filament for camera gear

Most of what gets printed for cameras is printed in PLA, and most of the time that is the right call. PLA is cheap, it prints without an enclosure, it holds dimensions better than anything else on the shelf, and for a lens cap that lives in a bag it will outlast the camera.

The trouble starts with the parts where it quietly fails, and it tends to fail in ways that are not obvious until the part is holding something expensive.

The temperature problem, which is worse than it sounds

PLA's glass transition sits around 55 to 60°C. Below that it behaves like a rigid plastic. Above it, it softens.

A car interior in direct summer sun reaches 60 to 70°C without much effort, and a black object sitting on a dashboard in that car gets hotter still. There are enough accounts of PLA parts drooping on dashboards that it should be treated as a certainty rather than a risk. If a printed part will ever be left in a car in summer, PLA is out. That covers a lot of photography, because gear lives in cars.

The second failure mode is slower and catches more people: creep. Under sustained load, at temperatures well below its glass transition, PLA deforms permanently over weeks. A printed quick-release plate with a camera hanging off it is under constant load. It will not snap. It will very gradually go out of true, and the first sign is usually that a clamp that used to hold now slips.

This is the single strongest argument for not printing load-bearing camera hardware in PLA, and it is why the Arca-Swiss plate project is worth printing in something else even though PLA would produce a crisper dovetail.

What each material is actually for

PETG is the sensible default for anything that leaves the house. Glass transition around 80°C, tougher than PLA, bends before it breaks, and it shrugs off the car problem. The costs are real but manageable: it strings more, it does not hold fine detail as crisply, and threads printed in it need a little more clearance because the surface is slightly rubbery. For grips, brackets, hoods, cages and holders, it is the material to reach for.

ASA is what you use when the part lives outdoors permanently. Around 100°C glass transition, and unlike ABS it is genuinely UV stable, so it does not chalk and go brittle after a season in the sun. The price is that it wants an enclosure, it warps if it does not get one, and it smells. For a tripod accessory that stays clamped to a fence, it is the correct answer. For a lens cap, it is overkill.

ABS is ASA without the UV resistance and usually a little cheaper. If the part sees daylight, use ASA instead. Its one advantage is that it can be smoothed with acetone vapour, which is genuinely useful for sealing a part against light and moisture.

TPU is the material people forget they own. It is flexible, in the 90A to 98A shore range for most spools, and it is the right choice for a whole category of camera parts: rear caps and body caps that need to grip without cracking, light seals, gaskets, strap padding, anything that clamps around a tube, and badges that have to sit flush on a curved body like the IR logo. It prints slowly, it prefers a direct drive extruder, and retraction settings that work for PLA will produce a mess. Budget an afternoon of tuning before you need it.

Nylon and nylon-carbon are the strong options. Untreated nylon is tough and abrasion resistant, which makes it good for parts that get repeatedly assembled and disassembled. It also absorbs water out of the air greedily, and a spool left out for a week will print badly until it has been dried. Carbon-filled nylon is stiffer and much more dimensionally stable, and it eats brass nozzles — fit a hardened one first.

Polycarbonate is stronger and more heat resistant than any of the above and is difficult enough to print that it is rarely worth it for camera hardware. If you are considering it, the part probably wants to be metal.

Choosing by part, not by material

The adapter row surprises people, since it puts the least durable material on the most precision-critical part. That is deliberate. Mount adapters need dimensional accuracy above all, they are threaded against metal, and PLA is the most predictable filament to hit a dimension with. They also live in a bag rather than on a dashboard. If your adapter will be exposed to heat, print PETG and expect to redo your clearance calibration.

The optical problem: black is not black

Anything that goes near the light path has a second requirement that has nothing to do with strength.

Filament is more translucent than it looks. Hold a printed part up to a phone torch in a dark room and a surprising number of "black" prints glow a deep red or grey. On a lens hood or a body cap this is a light leak, and on a body cap it is a light leak sitting directly in front of the sensor for however long the camera is in the bag.

Three things help. Print more walls — three or four rather than two — since most leakage is through thin sections. Use a filament that is actually loaded with carbon black rather than dyed, which usually means the cheap matte blacks perform better here than glossy ones. And check the finished part with a torch before trusting it.

The related problem is reflection. A printed lens hood with smooth internal walls will bounce light around inside itself and put veiling flare into your images, which defeats the point of the hood. The fix is texture: matte black paint, a flocking sheet from a telescope supplier stuck to the inner wall, or simply printing the interior surface with a coarser setting so it scatters rather than reflects.

Threads under load, and knowing when to stop

There is a line worth drawing clearly. A printed 1/4"-20 thread, cut into plastic, holding a camera body over concrete, is not a good idea. Plastic threads strip, and they strip suddenly, without warning.

For anything that carries weight, put metal in the load path. A heat-set brass insert, a captive nut in a pocket, or a hole sized for a real bolt to pass through. All three are trivial to design in, and the cold shoe adapter is a good example of the pattern: plastic does the shaping, metal does the holding.

The other honest limit is that printed parts have a grain. They are strong along a layer and weak between layers, and a part loaded across its layer lines will fail at a fraction of the strength you might expect from the material's datasheet. Orient the part so the load runs along the layers rather than trying to peel them apart. Where that is impossible, the answer is more walls rather than more infill, because in most camera-sized parts the walls carry nearly all of the load.

A note on drying, which is not optional

PETG, nylon and TPU all absorb moisture from the air, and all three print noticeably worse when wet: stringing, popping sounds from the nozzle, weak layer bonding, rough surfaces. Nylon is the extreme case and can go from usable to useless in a couple of humid days.

A filament dryer is the obvious fix. A cheap food dehydrator does the same job. Even a sealed box with a large bag of desiccant will keep a spool in reasonable condition between prints. If a material that used to print well suddenly does not, moisture is the first thing to check, before the nozzle and before the profile.