Product Design
Design notes: why this camera exists (and why it’s built the way it is)
This is the nerd page. If you just want to know what to buy, the Buying Guide is the practical one. This page is more about why the FTC-1 exists, how it evolved, and why certain design choices landed the way they did.
Why I ended up making my own camera
During COVID I got into 3D printing, which collided with a long-running photography habit and quickly spiralled into large-format film and perspective correction. There’s something deeply satisfying about perfectly straight verticals in architectural photos.
I was so hooked that in May 2023 I took my full 4×5 camera setup bike touring in Korea: folding field camera, three lenses, film holders crammed into a massive handlebar bag, tripod strapped to the back. It was ridiculous and good fun, but it also made some things very obvious: even a “light” 4×5 is hefty, setup takes time, colour 4×5 is expensive, scanning big sheets is a pain, and lenses with generous movements are priced accordingly.
Around then I was chatting with Andreas (@andreas_mass_design) about wanting perspective-corrected shots without all the overhead of large format. We both wanted something in medium format - more approachable, much cheaper to run, still with real movements. I immediately thought “I should make that”, opened Fusion 360 and started drawing.
What the camera actually is
The FTC-1 is basically a medium format technical camera. It takes 120 roll film using 6×7 backs (repurposed RB67 backs), uses large-format lenses in the ~47–135mm range, and gives 30mm rise, 5mm fall and 15mm of left/right shift. The back rotates 360° with magnetic detents and the ground glass is on a magnetic frame so you can swap between ground glass and film easily.
My aim was to sit between “no movements at all” and the high-end digital technical cameras built around 150MP backs that need micron tolerances and cost as much as a car. This is film-oriented and I’m not chasing what that kind of resolution demands, which hopefully means I can keep it at a price serious amateurs can justify without the stress of a studio system.
Prototyping
The first prototype was entirely 3D-printed in my shed. It leaked light in a few shots but, remarkably, it worked, and that was enough to keep going.
The second prototype added a CNC-cut timber handle and mostly eliminated the light leaks. I shot and developed a test roll the day before flying to Japan, decided it was “good enough”, and then used it for around 15 rolls during that trip. Very few issues — at that stage I thought I was basically done.
In Japan I met Oscar Oweson (@Panomicron) for coffee. I complained about tripod-mount stiffness. He showed me his latest prototype with CNC aluminium parts from China and, in essence, nudged me from “just add more plastic” to “maybe put some real metal in there”.
That turned out to be the inflection point: I went from printing everything myself to integrating CNC plates and 3D-printed metal for the bits that needed it, and the design stepped up a level.
Since then I’ve refined the rise mechanism, experimented with 3D-printed metal, and solved an embarrassing number of small issues. There’s still one tiny cosmetic error on my current prototype (which I’m not going to point out), but functionally it’s the camera I wanted.
Design and material choices
CNC aluminium
It might feel kind of stupid to “defend” metal in this application, but it does come at real cost financially and in terms of weight - and weight matters to me; I use this camera on a bike and while I don’t hike that much these days I’ve spent a fair amount of time with a very light backpack on my back (shout out to the PCT class of 2016).
My initial designs were all plastic, and it was pleasingly light and mechanically fine. However it had some serious issues with stiffness around the tripod mount - you’d move the camera and it would then sort of sway a bit before settling. This didn’t stop me using it and getting sharp shots but it took a lot more patience and I don’t think it was viable.
Initially, when I started exploring alloy at Oscar's suggestion, I was thinking about using “just enough” metal to get the stiffness I needed - something like a single plate through the middle - but I quickly went down the rabbit hole of using CNC more widely. I think this is nice and the solidness of the unit now is extremely pleasing and it remains “light” - but it’s still got a fair bit of metal on board.
I do often think about swinging back to my earlier intentions and paring back the metal to only where it is needed - a minimal version that has just a metal core and tripod plate with all other parts being nylon and going sans handle and cold shoe could be both cheaper and lighter. Not for today, but some time in the future.
3D printed plastic
While I said above I was drawn into the allure of CNC metal for a lot of the camera there have been several parts where that would clearly not make sense.
Lens cones are the obvious ones - they are reasonably voluminous with big areas of empty space. The cost of manufacture of these cones in a subtractive method would be significant - all my other CNC parts are cut from relatively thin plates with an acceptable amount of lost material. I can only assume cutting one of those cones would lead to a big pile of shavings and I don’t like that environmentally.
Conversely, printing them with an additive method means they can be far more efficiently created and still have all of the stiffness needed in these parts. Likewise, big parts of the rear standard and rotating back don’t benefit from jacked-up stiffness but are much lighter in plastic, so I’ve stuck with it.
All of my prototypes I’ve printed at home using FDM printing technology, which is perfectly fine and could be used for production - however, I only have one printer and no desire to build up a print farm so I am exploring other options for scaling. I may just outsource to a local print farm but I am more likely to move to MJF nylon printing as it gives a nicer finish. It’s actually a bit heavier as the plastic needs to print solid instead of having an internal honeycomb on the blockier parts, but I think it’s worthwhile as nylon in general is a more robust plastic than the ones you can sensibly print with FDM.
Maybe as I grow to bigger production runs I will also explore vacuum casting, which allows per-unit cost to drop but you do need to order quite a few to make it make sense, and it’s also not ideal for all forms so I may need to do some redesign to make the parts compliant.
Why so much information about plastic? I want to be clear that the cameras I ship will very probably not look exactly like the product shots (if you zoom in 500×) and may even change from run to run if I sample different techniques. However, they will always be black and dimensionally identical.
True to my word here - you can see I have struck out the comment above on MFJ - I was doing some prototyping and had some warping on thin areas that print fine in carbon-filled plastics at home. I can redesign or I can get the MJF supplier to print at funny angles - but then you can get a not great surface finish. I will stick to FDM printed at home for Run #1, and likely buy more printers for future runs.. Also it lets me print cones to order and if someone orders the wrong cone I can run them off a new one and get it in the post almost immediately, vs needing to wait for an order from an overseas MJF supplier.
Timber handles
In a camera where the other materials are somewhat severe and stark I really just want the timber to bring in some softness and character. But beyond that immediate look it’s nice to have a part of the camera that definitely breaks in instead of wearing out (not that I expect any of the other bits to wear out).
When I first put the current timber on I thought it was too blonde and I wanted something darker, but after using it intensively (especially on a sweaty-handed summer Japan trip) it’s gained a deep golden hue, more shine and really ties the camera together.
What of the shape and ergonomics - I did try a CNC-carved one on an early prototype (before I was CNCing any metal even) but what I learned through many, many iterations of 3D-printed forms and then committing to one in timber was it is very hard to make something comfortable for everyone (or even me on every day).
I was buying a knife in Japan and had the thought of “if this simple round wooden handle is comfortable here why wouldn’t it be for a camera?”. Initially I pivoted to a handle that was “pill shaped” with timber sandwiching plastic and metal but honestly this didn’t seem to gain me much comfort over the perfectly round pieces of dowel I was using to fabricate it - so I adopted that shape and honestly just haven’t looked back.
3D printed metal
The progression of the handle really led me to printed metal - once I had landed on wanting to hold the dowel captive I started thinking about how to do that and I worked out pretty quickly that a ring-like shape would be best but it ends up being a really awkward shape to CNC, and therefore costly.
However this was just as my main fabrication supplier started doing 3D-printed steel and I jumped right in - it ends up being considerably cheaper to print a part like this than machine it, even from very high quality metals. I started with stainless steel but have moved to titanium as it’s become more available, and this actually saves a not inconsiderable amount of weight for the cold shoe.
Acrylic ground glass
For now I’m sticking with an acrylic ground glass with fresnel - I like acrylic, it’s really flat at these sizes, doesn’t break when you inevitably drop it, and can be more optically clear than glass.
The penalty is the grind tends to have a bigger grain size than glass so it’s sometimes a slightly grainy image when looking at it through a powerful loupe. This would probably bother me more if this camera was being used for portraits or macro where extremely fine focus is required, but to be honest I hardly ever even use a loupe with this camera - my focus workflow is usually:
Is that building / tree / big object in the middle distance pretty well in focus?
Ok sweet, stop down to f/16 / f/22 and shoot.
Having said all that, one day in the future I may procure glass pieces if people want them, but for now I’d rather keep it a bit cheaper and use acrylic.
TR8 lead screw (8mm diameter) & anti-backlash nut
I know this probably looks like a seriously overbuilt component, and I did resist it for a long time. My earlier designs didn’t use a lead screw at all - they had rack and pinion gears as I could make them extremely lightweight, and they worked really well in terms of precision, but they needed a separate locking screw to make sure they didn’t slip at all. I frequently forgot to do it up and they didn’t slip, but it was still fiddly.
I moved to a lead screw about a year ago and initially used TR4 and TR5 trying to keep the weight down. The geometry of lead screws is such that they will only be reliably self-locking at a lead (how much the nut moves in a complete rotation of the screw) which is less than about 60% of the diameter.
On TR5 it was reliably locking with a 2mm lead, but got questionable with 4mm lead. The 2mm lead felt way too slow to adjust - I was spinning away like crazy to move the rise. I decided 4mm lead was the right balance of control and speed, so I bumped up to a diameter where that is getting a definitive lock.
I could possibly have moved to TR6, but a bigger diameter will feel smoother for a given lead (lower angle of the threads), and the other benefit of going up to TR8 is they’re so commonly used on 3D printers that parts like anti-backlash nuts are readily available without needing to get one machined, which saves manufacturing cost.
An anti-backlash nut (the brass bit with the spring) is basically a nut in two parts. The spring pushes them apart, which presses each part into opposite threads on the screw. The benefit is the nut always stays in contact with the screw, so when you rotate the screw it engages straight away rather than needing a little bit of movement first (the “backlash” we are trying to avoid). The penalty is slightly more friction, but for this application it’s worth it - it makes the rise movement feel far more immediate and reliable.
One reddit commenter told me it’s “overkill”, but I’m pretty confident it’s not. Also worth noting: the screw rides in sealed cartridge bearings I don’t expect to ever need to be replaced, but if they did wear out they are standard parts.
A note on lubrication; stainless steel and brass move really nicely with each other and to date I've never bothered to lubricate this part and have felt pretty good about the smoothness of the motion. However, moving to production I've been wondering about whether I should go better so I started considering options to apply some "factory lube" - PTFE spray goes on dry and lasts quite well so it's the obvious choice, but I had some misgiving about applying teflon given the current global concern around PFAS.. however the legends at WD-40 have recently released a PFAS free dry lubricant to address these concerns, so that's what I'll be using. I expect it to last a couple of years before reapplication is required, and it should be available at any good hardware store when needed in a spray can that will probably last the rest of your life.
Features and movements
Shift
Shift is the raison d’être of this camera — but why does it matter? When you’re taking a photo of a building (tall trees also benefit) you really want the film plane to be vertical, because that will mean the verticals of the building are also vertical.
But what if it’s a really tall building? If the camera is vertical you’re probably capturing a lot of ground, and not all of the building. Most people solve that by tilting the camera up — but then the top of the building is further away from the camera than the bottom is. Lines converge toward the top and you lose the clarity of form.
I've done the mock up below to show this - lets say you stumble upon this frankly gorgeous brutalist building (thanks GPT), so you setup your camera on the tripod with a good focal length for the shot, if the camera is level and the horizon line is in the middle of the frame you'll get something like the shot on the bottom left, too much ground and the building is cropped. If you respond to this like the average iPhone shooter you'll angle the camera up to look at the building, which gets it in the frame, but by changing the angle you've also caused a subtle but significant effect where the top of the building is now further from the film / image plane that the bottom and as a result the vertical lines of the building converge. The stately beauty of the building is lost - see bottom middle image. Shift can fix this by letting you maintain a level camera but adjust the vertical perspective as required, nailing the bottom right shot.
To further understand how this works, think of the film as cropping from inside the large image circle of a large-format lens: the lens “sees” a huge field of view and by shifting you’re moving which part of it your film sees. You get the top of the building without tilting the camera, and you keep those square, sharp verticals.
Horizontal shift is similar. There’s often a best spot for a scene — maybe it makes a facade square, lets you look through a doorway, or gives the best angle. But what if from that ideal perspective the subject isn’t where you want it in the frame? Horizontal shift lets you stand in the right place and then control the frame from there.
The below diagram shows a bit more of this image circle concept and how the camera "selects" from within the image circle. 
Why no tilt mechanism?
Tilt is for angling the focal plane relative to the film plane (Scheimpflug). It’s useful for macro and tabletop work, and sometimes landscape (foreground flower and background mountain both sharp, etc). It can also be used for the opposite effect and make things look miniature. You do you.
But in architecture it’s a risky business — you don’t want weird outcomes like the top of a building being sharp when the bottom is out of focus. I’ve thought quite a bit about mechanisms that would let me build tilt in, but because I use helicoids to focus the lenses it can’t ever be axial, and it all gets a bit weird for focus control. More than that, I didn’t like the extra degrees of freedom: I worry that when I wanted zero tilt (95% of the time) I wouldn’t have exactly zero.
There are times I do want tilt, so the approach that makes the most sense to me is offering extra lens cones with fixed tilt (calculated based on focal length, ranging from 2° to 6°). These won’t cut it for serious macro work (please don’t buy this camera thinking it’s built for that), but they will allow enough Scheimpflug in landscape work for that “everything in focus” look.
Extra cones are easy to produce and give a very confident amount of tilt without any possibility of wandering. They’re a little bulky but not heavy to have in your kit, and changing a cone over is about a minute in the field.
I've been testing a 2.5° tilt cone with a 65mm lens on the beach and it's been great for that kind of work - it looks funny like the cone melted a bit, but I think it's cute. It's probably not super apparent from the below shot at web resolution, but the foreground rock is every bit as sharp as the massive rock in the background, and this kind of deep landscape is exactly what the tilt cones help achieve.
I think there's a fair chance if your architectural work skews lower rise (suburban houses and their lawns for instance) that the tilt cone might be an "always" proposition, but I haven't done enough testing yet to qualify that so I'm still selling a no-tilt cone as the default that ships with each body. My experience with how I shoot architecture is f16 and a sensible hyperfocal focusing distance gives me all the depth of field I need (frankly I usually don't remember / bother to refocus between shots once I've got it locked in).
What about digital?
This is sort of a funny one as I have a bit of cognitive dissonance about it — I designed the camera around 6×7 and have never deviated from that core dimension, but I actually use mine predominantly with digital these days.
Quite simply I travel a fair bit and mostly take a big volume of photos while travelling and I got completely sick of the stress of x-ray machines, so I sought out an old P45+ (mine was made in 2008) which I got for a steal from a slightly dodgy looking eBay listing.
I got a Phase One mount working and have been using that back for the past year and a bit. It’s really fun as it has no live view, so I still compose on a ground glass (just a smaller one), and while it can give me a review afterwards the screen is basically a Nokia candybar circa mid-2000s and it’s not worth much — I get a kind of “framing looks about right and the histogram says it’s not blown” confirmation and keep moving (after I’ve taken an LCC frame and other fun foibles).
Despite the fact it works great for me, I don’t really have any intention of selling a digital version. The majority of Phase One backs are way higher resolution than mine and while I have no real indication (based on my 39MP images) that the tolerances of my design and manufacturing wouldn’t hold up, I don’t have a 60, 100 or 150MP back to test with and I don’t want to let anyone down. This design is toleranced around film, and I have pushed it past the limits of film considerably with my current back — I’m not warrantying it should be pushed past that.
Also, the film plane is located very slightly differently to the RB67 mount version, so I’d need different lens cones to compensate etc, and I’m not offering all that mess as a standard product any time soon. If a bunch of P25 and P45 users come out of the woodwork maybe I’ll do a CCD-lover run later.