FFD on my Completely 3d Printed 16mm Film Camera - Looking for advice

[Henry kidman]

Hey guys, 

Firstly, I have only shot on film once, and I don’t own a camera, so forgive me if my terminology is not perfect! 

 

Im sorry if this is the wrong place to post this. im in the process of designing and building my own 16mm film camera. I have the transport reasonably sorted (ignore the take up spool spinning in the wrong direction and not at the correct speed, still need to code that part out), however I still want to refine the pressure plate and add a spring on the rail for lateral support. 

 

However, at the moment I’m trying to dial in the flange focal distance. I’m using a Sony e mount on the camera so it has a very short FFD of 18mm. I understand that the depth should be set a slightly short of that so it ends up focusing just past the emulsion. After reading, I’m thinking I get a depth micrometer and put a gauge block against the gate then take the measurement that way. Is a depth micrometer that does 0.01 mm fine enough or do I need to be investing in a 0.001mm? Is the depth micrometer the best way? How do actual camera techs do it?

 

Also if you have any other tips for the camera please feel free to let me know!! I will link a little video of it running - 

 

Thank you in advance:) 

Henry

 

[Tyler Purcell]

So 3D printed material, no matter what it is, will not have the tolerances for proper FFD, never going to happen. Manufacturers spend an ungodly amount of time engineering the gate to flange distance and making it perfect. The gate and lens mount, really need to all be made of metal and somehow connected. 

So the tolerances are based directly on film channel/float. Some cameras have a pretty large film channel, this is the gap between the pressure plate and aperture plate. Metal gates are used because you can polish the side rails and the film can be at the same depth as the physical aperture. The pressure plate, shouldn't actually be putting pressure on the film itself, it should be very smooth but held firmly in place by the gate laterally to prevent wobble. 

Yea, you will need gauges that go down to .01, as the FFD range is .00-.03mm, depending on how much float the film has. Some cameras can be run at .03 and not be a big deal, others need to be spot on .00 and that's the tricky part. Engineering the pulldown, the consistency with the FFD across the frame, film channel, a spring loaded side rail gate, timing with the shutter, all of these things are very challenging to get right and I'm afraid, no way a 3D printed camera would actually create proper images. Someone tried it with a 2 perf 35mm 3D printed camera and it was basically unusable. The tolerances are incredible, .01mm off and you go from working to not working. As someone who services film cameras for a living, it's a miracle any motion picture camera works at all. 

I commend the work tho, I'm glad to see you playing around with it. I'd love to see if any images come out. 

 

[Henry kidman]

Hey Tyler, 

thanks so much for your input. Yeah the timing of the shutter with the film movement mechanism was a little bit of a headache, but it now works in such a way that whilst they aren’t mechanically linked as each has its own stepper motor, the film advance sprocket waits until the motor has completed a specific amount of its rotation before it will move, this way it’s not relaying on perfect time keeping or complex mechanical links. 
 

yes I did fear the ffd would be the part of this that stumps me. With that being said, I have ordered a depth micromiter so I’m just going to try to get it dialled as best I can. If you were servicing a camera and the ffd was out, how would you adjust it? I’m sure that varies from camera to camera but is it a shimming set up, or are you adjusting set screws? 
 

and as for the pressure plate its interesting you say it’s not meant to apply any pressure on the film it’s self. So should I be making a tiny shelf that the pressure plate pushes against that is 0.15mm above the height of the gate so the film can just fit through the gap? 
 

One final question, if I were to purchase a piece of ground glass and place it on the film gate to be able to visually check focus while I’m adjusting the ffd, would the result be indicative of the result when using film as far as focus goes? 

[Aapo Lettinen]

1 hour ago, Henry kidman said:

Hey Tyler, 

thanks so much for your input. Yeah the timing of the shutter with the film movement mechanism was a little bit of a headache, but it now works in such a way that whilst they aren’t mechanically linked as each has its own stepper motor, the film advance sprocket waits until the motor has completed a specific amount of its rotation before it will move, this way it’s not relaying on perfect time keeping or complex mechanical links. 
 

yes I did fear the ffd would be the part of this that stumps me. With that being said, I have ordered a depth micromiter so I’m just going to try to get it dialled as best I can. If you were servicing a camera and the ffd was out, how would you adjust it? I’m sure that varies from camera to camera but is it a shimming set up, or are you adjusting set screws? 
 

and as for the pressure plate its interesting you say it’s not meant to apply any pressure on the film it’s self. So should I be making a tiny shelf that the pressure plate pushes against that is 0.15mm above the height of the gate so the film can just fit through the gap? 
 

One final question, if I were to purchase a piece of ground glass and place it on the film gate to be able to visually check focus while I’m adjusting the ffd, would the result be indicative of the result when using film as far as focus goes? 

There has been lots of discussion here on the forums about the limitations of materials and 3d printing and electronic vs mechanical links. 

3d printing is just not a good technology to make every and all parts of a movie camera, and electronic links are not good idea if one needs high precision, high speed and reliability at the same time. 

3d printing out of metal like stainless steel is possible but cnc machining is cheaper on most type of parts with potentially better end results. Compared to 3d printed plastic, ANY old camera including the cheapest Soviet models has better quality film transport and gate than a 3d printed plastic one would be.

Fully 3d printing a camera out of plastic does not make sense, it will be a sub par product whatever you do and any 50 bucks old camera will blow it out of the water image quality wise. I think you have confined yourself inside a "I must make this entirely with 3d printing" box which limits your engineering process too much and you will never make a good working camera for actual filming unless changing some goals here. Allow yourself to use proper metal for dimensionally critical parts and a mechanical link for movement for example. You will get much better camera that way. Stainless steel, aluminium, brass, spring steel in selected critical applications. Then rest can be 3d printed if necessary. But allow yourself to consider on every single component: does this need to be metal? Can it be printed or is cnc machining and polishing and coating better option? How much each option affects the price and reliability? 

I would highly recommend getting some old movie cameras for evaluation how they did things. And if you find a good working used part just scavenging it from old camera instead of making basic stuff like rollers and sprocket wheels and gears from scratch 🙂

[Henry kidman]

Hey Aapo,

Yes very good points. And I would agree, cameras made by companies with budget, knowledge, talent and machines that far surpass my own will obviously be better, however that isn’t really the point. I want to make one my self. A modern digital camera is objectively cleaner and more practical in every way than a 16mm camera, and yet here you are. There is a charm about the process and the imperfections.

the 3d printing constraint isn’t some arbitrary challenge I have given my self, it’s out of necessity. I sadly don’t have access to a cnc machine, or the plethora of required supporting tools and knowledge, but I do have a 3d printer. Almost everything in the camera would benefit from being metal, but I’m curious how far I can make it with what I have.
 

Borrowing parts from old cameras is a good idea, specifically parts like the sprockets where dimensions are very important, however, I have done the work to get these sprockets of my own to fit properly so I’m pretty happy with them now!  And once again, I’m stubborn, I want to do it my self!! 
 

I would also have to disagree with electrical connections being less accurate than mechanical connections. We are constantly seeing mechanical links being swapped for electrical, think aircraft flight controls, throttles, electronic digital shutters. they offer many advantages, cost saving, accuracy and most importantly simplicity - fewer moving parts means less compounding play through small tolerance inaccuracy's.

I also think 3d printing might surprise you now, it has come a long way, and whilst far from having the dimensional properties of metal, it’s pretty cool to see how affordable it has become to be able to print things with a resolution of 0.017 mm. 
 

if you have any tips on measuring FFD, let me know!! 
 

cheers, 

henry 

[Aapo Lettinen]

1 hour ago, Henry kidman said:

the 3d printing constraint isn’t some arbitrary challenge I have given my self, it’s out of necessity. I sadly don’t have access to a cnc machine, or the plethora of required supporting tools and knowledge, but I do have a 3d printer. Almost everything in the camera would benefit from being metal, but I’m curious how far I can make it with what I have.
-----

if you have any tips on measuring FFD, let me know!!

Most of other people don't have personal access to high quality cnc machines either. We design the parts so that the first test design can be 3d printed to see if it is approximately correct design. Then when the design is finessed enough, it is time to look for a outside company to do a test piece from the file for reasonable price. When the test piece arrives and has been thoroughly evaluated, it is possible to decide if the design needs some final touches or if it is already possible to order the final parts from the outside company, machined to final tolerances with final surface finish.

it is not necessary to purchase a 300k+ priced cnc machine when one can hire a company to make the part for you from the step file. Just design the part so that it is possible to cnc machine and then quote from various places how much it would cost to machine it out of aluminium or steel etc.

you can check for example pcbway and jlcpcb for cnc quotes if can't get good quotes from AU.

 

the ffd needs to just be calculated close enough when designing parts, then manually tuning when fitting. and when almost there, sand down and polish the shim until you get it perfect. you may need to make your own ffd gauge if there is no readily made available. the gauge would be a good project to test outsourced cnc machining, you need to make it out of metal. get a suitable dial gauge and design the rest of the parts by yourself

 

[Aapo Lettinen]

Here is an  example of stuff I made last year. the first test pieces are 3d printed to save costs, then when the design is final I will check various places how much they charge for final one made. The first print is home printed from pla and not sturdy enough for use but works for optical testing. The final part is cnc machined from stainless steel by pcbway with basic surface finish. the final part cost a little over 100usd with shipping costs to make which is why I had to be sure the dimensions were at least close to correct ones, thus made multiple home 3d prints to finesse it (prints cost maybe from 1 to 3 usd a piece so one saves a lot if can find basic errors with printing before ordering 100usd+ final parts made out of metal)

The camera handle made the same way btw, it is cnc machined aluminum. I was so sure about the design that did not bother to 3d print a test piece. the handle has 1/4" and 3/8" threads for mounting accessories. I think it was something around 100usd with shipping costs, made by jlcpcb. black anodizing too

Edited April 21, 2025 by Aapo Lettinen

[Aapo Lettinen]

personally I would make the gate from stainless steel, the pulldown claw spring steel, pulldown axles some kind of steel which is available, the pressure plate aluminium, the shutter aluminium. All of these can be cnc machined.

then the main support structure connecting the gate and pulldown+shutter assembly to the lens mount would be aluminium on basic designs and stainless steel on very high quality designs. I would use cnc machined aluminium on this one for the much lower costs. Basically it would be a kind of U-shaped structure which seamlessly connects the gate and lens mount together to control the ffd precisely. a shim under the lens mount is used to fine tune the distance. you can get the shim cnc made too though would be cheaper to use some available Chinese shims from lens adapters and just design the mount so that they can be used.

film transport depends, the rollers are perfectly fine if plastic. the sprocket wheel depends but often it can be plastic if the sprockets themselves are metal. for example a metal ring with sprockets which is attached on a plastic roller to form a full sprocket assembly.

All the other parts of the body which don't have high durability and / or dimensional demands are made out of plastic. on low quantity it is economic to 3d print. if making like 20 cameras it is often more cost effective and higher quality to cnc machine the plastic body too if the shape is such that it is possible to machine

[Henry kidman]

That is very impressive! 
Who knows, it may come to the point that I need to send parts off to be done properly, however, the goals for this camera are very modest, it’s not trying to be better, or more practical then any of the other 16mm cameras available, it doesn’t have a view finder and it’s missing a lot of the other niceties. I just want it to take a nice clean image, and for it to be able to reasonably accurately focus. 
 

so we will see when I run some film through how far off I am from that. Depending on how that goes I might have to step off my high horse and get the parts machined!! 
 

thanks for the examples, they look very professional 

[Aapo Lettinen]

4 minutes ago, Henry kidman said:

That is very impressive! 
Who knows, it may come to the point that I need to send parts off to be done properly, however, the goals for this camera are very modest, it’s not trying to be better, or more practical then any of the other 16mm cameras available, it doesn’t have a view finder and it’s missing a lot of the other niceties. I just want it to take a nice clean image, and for it to be able to reasonably accurately focus. 
 

so we will see when I run some film through how far off I am from that. Depending on how that goes I might have to step off my high horse and get the parts machined!! 
 

thanks for the examples, they look very professional 

thanks! yes it is often better to machine stuff for final parts but 3d printing is very good technique for first prototypes and saves lots of costs on prototypes. You can work very efficiently by replacing one part at a time with cnc machined metal version when you are sure that part is final. I would first replace the gate and pulldown with metal ones so that you have possibility to get higher quality images to begin with. then can replace other parts later on one at a time when feeling necessary and have collected resources for it 🙂

cnc machining has some limitations with sharp corners and one may need to have additional finishing ordered for parts which for example need sharp cornered hole cut. For example it is possible to machine the film gate with rounded corners on the gate opening but if needing to have as sharp corners as possible one may need to order additional wire saw or EDM finishing to get them as small radius as possible. machining cavities has some limitations too.  For the gate I would use stainless steel or chromed brass if steel is too expensive. the surface needs to be smooth and polished on rails but other parts are possible to leave a bit rough if they are dimensionally accurate. the Soviets only polished the parts which touched film surface and other as left unfinished... most Western cameras are finished all the way through.

with cnc and metal it is easy to get high quality threads on the parts which is often important when assembling stuff.

top:  Eclair NPR 

bottom: Kinor 16cx-2m

[Dom Jaeger]

Hi Henry, I’m a Melbourne based cine camera and lens technician, roughly 30 years experience. There are not many movie cameras I haven’t worked on over the years, though some I’m much more familiar with than others. Great to see Aussie ingenuity and curiosity at work, good on you! 

While I agree with other comments here that 3D printing isn’t really suitable for the critical parts of a movie camera, if you’re just experimenting and playing around with what you have, then go for it. Nobody invented anything by listening to people who told them something can’t be done. 

To measure flange depth we use a depth gauge with 0.01mm accuracy, along with a backing block that is placed in the gate. We also use a cylinder blank accurately machined to the nominal flange depth in order to zero the depth gauge. To adjust flange depth, we use shims of various thicknesses. They can be metal or plastic, so you could cut your own from plastic shim stock. 

While 35mm cine cameras often utilise a very precise film channel with a pressure plate that doesn’t press on the film, the pressure plates on most 16mm cameras actually do exert pressure. That might make it a bit easier to manufacture, though you need to get the pressure right. I would suggest looking at something like a Bolex, which uses a very simple mechanism that still produces quite stable images. 

I have manufactured my own S16 Bolex gates out of steel, which could possibly be of use to you. Or you could get some bits out of an old junker Bolex. You do need to make sure the lens mount is quite parallel to the gate, ideally within 0.01mm across the diameter. The flange depth should also ideally be within that tolerance, though all this accuracy is somewhat contingent on other factors, including your expectations. I don’t want to go down that rabbit hole yet however.. the question I have for you is, if you don’t have a viewfinder system, how do you plan to focus a lens? There have been a few systems invented over the years, the most common one before reflex viewing was a rack-over system, where the lens mount is shifted over to a ground glass and viewfinder for focussing, then shifted back to the gate for filming. 

If you ever come to Melbourne, I’d be happy to show you my service department and some of the tools and techniques we use for camera checks. 
 

 

[Tyler Purcell]

13 hours ago, Henry kidman said:

Hey Tyler, 

thanks so much for your input. Yeah the timing of the shutter with the film movement mechanism was a little bit of a headache, but it now works in such a way that whilst they aren’t mechanically linked as each has its own stepper motor, the film advance sprocket waits until the motor has completed a specific amount of its rotation before it will move, this way it’s not relaying on perfect time keeping or complex mechanical links. 
 

Using electronics to join multiple motors together is fine, but with 16mm cameras especially, you can't use a sprocket intermittent to pull the film down, you will need some sort of pulldown claw. A sprocket, won't work. The reason cameras use pulldown claws, is because they can have very tight tolerances. A sprocket has very bad tolerances, which would cause lots of up and down movement in the image. Even if you were to reduce that with a spring loaded side rail, it would still be an extremely unusable/unstable image. I know it runs through the camera ok, but those micro movements can't be seen by the naked eye. 

As Aapo suggested, the best thing to do is start with a movement from another camera. So that all your main components are made of metal already. I really like the ACL movement (minus the mirror) as it's mounted to a nice thick piece of metal, that you could easily find a front housing for, even perhaps an ACL one. To save even more money, re-working a K3 movement, is probably pretty straight forward. 

13 hours ago, Henry kidman said:

yes I did fear the ffd would be the part of this that stumps me. With that being said, I have ordered a depth micromiter so I’m just going to try to get it dialled as best I can. If you were servicing a camera and the ffd was out, how would you adjust it? I’m sure that varies from camera to camera but is it a shimming set up, or are you adjusting set screws? 

Shims yea, that's the way most cameras work. Arri has a few cameras that the entire movement can move back and forward but for the most part on 16mm cameras, it's done with shims between the lens mount and the body. 

13 hours ago, Henry kidman said:

and as for the pressure plate its interesting you say it’s not meant to apply any pressure on the film it’s self. So should I be making a tiny shelf that the pressure plate pushes against that is 0.15mm above the height of the gate so the film can just fit through the gap? 

The kit to measure would have a flat piece of metal for measuring that is the width of the 16mm gate. A FFD gauge, which usually has a flange that you would push onto the lens mount. Then a tester tool, so you can set the FFD gauge to zero before putting it onto the camera. FFD tools aren't horribly expensive and are available for sale, but the 16mm width flat piece of metal, maybe harder to find. I had to buy one from another tech. 

13 hours ago, Henry kidman said:

One final question, if I were to purchase a piece of ground glass and place it on the film gate to be able to visually check focus while I’m adjusting the ffd, would the result be indicative of the result when using film as far as focus goes? 

You can't use the ground glass method because it adds too much depth. A very thin piece of smoked paper can work on 35mm cameras to see if you're in the ballpark, because the image is so much bigger, but on 16mm cameras with that small of an image, it's just impossible with out proper FFD tools. No matter what, to "dial it in" you will 100% need the tools, there is nothing you can do about that. A collimator could be used theoretically, but they're more expensive. 

[Aapo Lettinen]

It is entirely possible to make a simple pulldown system from scratch. simplest cameras like Krasnogorsks use this style of very simple movement which should be pretty "easy" to make to work OK.

there is a cam driving the end of the pulldown claw up-down-towards gate -away from gate. The claw has a guide pin in the middle-ish which slides on guide rails, translating the movement from the cam to the desired type of "pull-the-film-down, retract and go back up, repeat" type of movement.

It is not perfect but it is simple and relatively easy to make to work well enough for home use. needs to be made out of proper metal though, no plastic can tolerate this use.

I quickly modeled some reference images, nothing is in scale and you may want another set of guide rails on top to add stability but the basic principle is easy to see.

 

[Henry kidman]

10 hours ago, Dom Jaeger said:

Hi Henry, I’m a Melbourne based cine camera and lens technician, roughly 30 years experience. There are not many movie cameras I haven’t worked on over the years, though some I’m much more familiar with than others. Great to see Aussie ingenuity and curiosity at work, good on you! 

While I agree with other comments here that 3D printing isn’t really suitable for the critical parts of a movie camera, if you’re just experimenting and playing around with what you have, then go for it. Nobody invented anything by listening to people who told them something can’t be done. 

To measure flange depth we use a depth gauge with 0.01mm accuracy, along with a backing block that is placed in the gate. We also use a cylinder blank accurately machined to the nominal flange depth in order to zero the depth gauge. To adjust flange depth, we use shims of various thicknesses. They can be metal or plastic, so you could cut your own from plastic shim stock. 

While 35mm cine cameras often utilise a very precise film channel with a pressure plate that doesn’t press on the film, the pressure plates on most 16mm cameras actually do exert pressure. That might make it a bit easier to manufacture, though you need to get the pressure right. I would suggest looking at something like a Bolex, which uses a very simple mechanism that still produces quite stable images. 

I have manufactured my own S16 Bolex gates out of steel, which could possibly be of use to you. Or you could get some bits out of an old junker Bolex. You do need to make sure the lens mount is quite parallel to the gate, ideally within 0.01mm across the diameter. The flange depth should also ideally be within that tolerance, though all this accuracy is somewhat contingent on other factors, including your expectations. I don’t want to go down that rabbit hole yet however.. the question I have for you is, if you don’t have a viewfinder system, how do you plan to focus a lens? There have been a few systems invented over the years, the most common one before reflex viewing was a rack-over system, where the lens mount is shifted over to a ground glass and viewfinder for focussing, then shifted back to the gate for filming. 

If you ever come to Melbourne, I’d be happy to show you my service department and some of the tools and techniques we use for camera checks. 
 

 

Hey Dom, 

Thank you so much for your response, and so cool that you are in Australia!

I just got my self a depth mic, a gauge block and a piece of granite for something flat so lets see how i go! Do you have a link to where i can find some blank plastic shim stock, I cant seem to find any,  just pieces that are already cut to shape.

Yes that was my understanding when i used the k3, that the pressure plate was pushing directly on the film. My current pressure plate set up is a big janky, however it does hold the film with just a little bit of force, so it may work. I'm using this sprung plunger design, which is a product of implementing it in the design last, and now having to work around things. Its biggest problem is that there is a little bit of lateral wobble in it which I'm sure is not ideal, if it is causing too much trouble i will do a proper re design.

Thats so cool that you have made gates out of steel, where do i go to have a look at these? did you machine them your self or have them sent off? 

Thats a good question about focusing. When i was doing research, many people were saying the view finders on the 16mm cameras (the more budget ones anyway) are not to be trusted for focus, so instead they suggested its best to use a tape measure and the markings on the lens. Whether this is true or not i don't know, however,  knowing that people do focus just using the marks on the lens gave me the idea to ditch the view finder as it was a lot of extra complexity.

I would love to have a look at your service department, I'm in Melbourne every couple of months for work so if you don't mind I might send you a message next time as it would be fascinating to see how you do it, and would definitely help me with my cam!!

Edited April 22, 2025 by Henry kidman
add photos

[Tyler Purcell]

6 hours ago, Aapo Lettinen said:

It is entirely possible to make a simple pulldown system from scratch. simplest cameras like Krasnogorsks use this style of very simple movement which should be pretty "easy" to make to work OK.

Yea, your drawings are basically similar to what I would do, even if it was 3D printed. One COULD theoretically just yank that assembly out of a K3, maybe cut the front of the K3 off and just power it by a motor, it maybe the easiest way to go about it. The nice big flange the movement is mounted to in the ACL, that's just grand. I really like that design and it's why I'm gonna be using it for my prototype. We still have to get the metal machines in, but with our current commander and fascist, there is probably no way we'll ever be able to afford them OR the raw materials anymore. We ordered a new style of 3D printer (at great expense) two months ago, but seeing as they're coming from China, I doubt we will see it in the next month or two. But theoretically it can print more accurate parts for prototyping that would allow us to make the thing 3D properly (with a copied movement) and then send it out for metal manufacturing. In the US however, there is very little metal prototyping going on, they only want contracts for big jobs. So we would have to send over to Asia, but now that we're cut off from them, I have no idea what to do. The entire project is sidelined. 

[Dom Jaeger]

Apart from a drawful of metal shim stock I've accumulated I've been using a book of Artus shim stock for about ten years, it goes from 0.01mm up to 0.75mm.. but I'm not sure where you'd get it now. This is part of the problem of trying to make something with very fine tolerances, you end up needing specialised tools and materials. There's this from RS Australia, but it only starts at 0.025mm and is rather expensive:

https://au.rs-online.com/web/p/shim-kits/0518778

Apart from ensuring the film is flat against the gate, a pressure plate also helps to keep it stable while the sprockets are still feeding the film through. The way movie cameras work is the sprockets are continuously driving the film, creating a fixed loop, while a pull-down mechanism creates the intermittent movement required for exposure and then moving to the next frame. If you don't have something like a registration pin you need another method to stabilise the film during exposure. Usually it's a combination of a pressure plate and a side spring (which is used to maintain lateral steadiness). If you're trying to use an intermittent sprocket drive to advance the film, I think you will find it both not accurate enough, and too aggressive on the film perfs. A small fixed loop is much less inertia to start and stop in such short time frames. But see how you go I guess.

I had my S16 gates made, they're quite accurate, but specific to the Bolex design. You could scavenge a standard 16 one from an old non-reflex Bolex pretty cheaply, and get the associated claw and side spring/lateral cage along with it. Or use parts from a K3  if you have one of those.

For focusing, if you rely only on lens marks and distance measurement then you need a few things to be right. You need the flange depth to be spot on, and you need the lens to be perfectly collimated. And if you want to use telephotos, then it becomes very hard to just use distance marks and still get a sharp image. It could work for wides and mediums, especially if you are stopped down a lot. If you are using stills lenses however, which are usually not very accurately marked with distances, and can have short focus throws, you might struggle to consistently get sharp focus. In the professional film industry where I work, crews do use distance marks, as well as focus aids and monitors, but the lenses are all thoroughly checked for mark accuracy, and they are proper cine lenses with long focus throws, many finely scaled distance marks and no internal backlash. 

You also have no way of knowing what you're framing without some kind of viewfinder. There have been different solutions to this problem used over the years, from a simple tube with framelines for different focal lengths, to more complex zooming viewfinders with parallax correction, to viewfinders with turrets and a selection of matching lenses. 

 

 

[Friedemann Wachsmuth]

I have no words for how much I love this from all angles.

 

[Brandon Paterno]

Yes, this is very exciting!!

Still seems to be major issues with registration. If the bar is so low, though, for a 3d printed camera that we should consider computer registration after a scan... can some overseas company make a replica of this movement for little low quality experiments? Well on the democratization of filmmaking... there are tons of mass produced 3d printer components in use here. Include both a feed and take-up spool and use a crappy machine vision lens. Maybe it was always possible to make something like this but is the expectation of computer stabilization an acceptable way to let people work with motion picture film material affordably? Personally I'm on the fence, because the footage made by this looks completely unusable outside of the computer. Good luck projecting the image produced by it. In some ways this is just a film filter machine too. 

[David Bernard Anderson III]

I think this post is a great litmus test for those who want to push the accessibility of film forward like @Aapo Lettinen and @Henry kidman do. There’s a lot of knowledge that’s left with the older generations as they transition, and there are a lot of people, including myself, who want to learn more and make film as accessible as it was in its golden era. Techniques change, but the message is the same. Love this! 

[Tyler Purcell]

6 hours ago, Brandon Paterno said:

Personally I'm on the fence, because the footage made by this looks completely unusable outside of the computer. Good luck projecting the image produced by it. In some ways this is just a film filter machine too. 

We're missing part of the story, partially due to the poor scan quality. As I said in another thread, this video kind of vindicated my concerns. It's far more complicated than just moving film. 

 

[Tyler Purcell]

3 hours ago, David Bernard Anderson III said:

I think this post is a great litmus test for those who want to push the accessibility of film forward like @Aapo Lettinen and @Henry kidman do. There’s a lot of knowledge that’s left with the older generations as they transition, and there are a lot of people, including myself, who want to learn more and make film as accessible as it was in its golden era. Techniques change, but the message is the same. Love this! 

Well, to be fair, there is nothing in this video that hasn't been done before. People have been 3D printing motion picture cameras for years. It's just, most people who would do such a thing are nerds and they aren't making YouTube videos, they're engineering stuff. We haven't seen an open source motion picture camera because where is the money? In an economy where everyone is selling their cameras to pay for food, I don't think there is a business model currently that makes sense. 

Aapo has done a great job making some cool motor/board combos for a select few cameras. However, he doesn't have the resources to make them forever. If you want to donate a million euro's to his cause, I'm certain he would be able to make many more cool products that potentially make an impact. 

I also have a camera design that we've been working on for years, but after talking to several vendors, I just can't see any future where it's financially viable. One off  "experiments" like Henry's aren't solutions either. A true solution will be a well funded company, with engineers and guarantees that they won't fold after the first batch is released like nearly all startups. The math I've done says, a $15k body with $5k magazines is possible, but what owner is willing to risk $30k on a "new" companies camera, even it were really nice? The answer is, potentially 15 - 20 people globally, but that's not enough for the investment NOR is it enough to keep inventory around, train service providers and keep the doors open. We saw this happen with Logmar, who did build an unexpectedly good super 8 camera years ago, only made 50 I believe and haven't really been able to springboard that success into anything else that's readily available. 

One thing most people didn't notice in the video is that the 3D printed camera is covered in metal film. This is because PLA and SLA materials (the two main ones used on FDM printers) deform under the sun. So yea, my entire commend about focal distance which is what led to all of this conversation in the first place, still holds water. Even if you can make it "good enough" to create an image, you certainly can't keep it that way for long unless it was stored full time in a cool, extremely dry place for it's entire life. So yea, you have to make cameras out of metal, at least the mid plane where the movement, gate, shutter, pressure plate and lens mount attaches. Sprockets also need to be metal, because they will scratch otherwise. All of those pieces are easy to get out of broken/damaged cameras tho. So if you wanted to make a "prototype" out of 3D materials, you could just source those parts from a pre-existing cameras that don't work. I keep telling the people the K2/K3 is a goldmine in this way, but very few listen. Then all you do is re-house the thing is a 3D printed housing with modern home-made motor controllers and a video assist. I mean it's been done already, it's called a K4 and it's all metal, available for purchase any day of the week for $5k.

If you ever want to learn about this stuff, we have workshops all the time and are open for customers whenever you want to visit pretty much 7 days a week in Los Angeles.  I'm one of a small group of independent motion picture camera service technicians who keep these cameras running, but I take it one step further and offer education opportunities as well on a regular basis, including knowledge dumps on YouTube. I think getting as much information out there as possible, is a great way to record the broad brush strokes. However, the trick with techs isn't the big brush strokes, most people can do those things. In the past, techs had some basic tooling to machine components back into spec, but for the most part when things were broken, there were always parts available. We've reached a period in time where parts, even for the gray body Arri's, may stop existing for the first time in the history of modern motion picture cameras. So tech's are no longer parts replacers, they are having to manufacture new parts from scratch which is costly AND takes away from other duties. For instance, I had to stop everything I was doing last week to manufacture a single part from scratch on our lathe and mill. Took us around 18 hours to make, but I can't charge for that time, the customer would have never bought the part. So instead of taking 18hrs and using it to fix cameras, I had to basically give all the time away to help a customer in need and we did just that. My fear is as time goes on, it will be more and more challenging to offer lower-cost services like that and people will simply not get their cameras fixed anymore. Just in the last few weeks, I've quoted people for some repairs that they couldn't afford. It's a real shame because they have 100% broken cameras, unusable at all, so putting the money into them, at least gives them value. Anyway, it's a real challenge from my perspective and quite a few cameras we get, offer up new challenges (in terms of failed parts) that really require quite an investment in time. Some people can afford it, some people can't and I'm seeing the ladder become more and more common. 

[David Bernard Anderson III]

[Aapo Lettinen]

I think one really needs to decide first whether one wants to make working film cameras for actual filming in reasonable time, OR if one is more interested in making technical research and prototypes spending lots more money and time and potentially never getting to actually make real movie projects with the devices, only small tests (with potentially disappointing results compared to the enormous resources used).

I am way too practical person to spend years making something out of scratch when I can just take a truckload of old cameras cheaply and scavenge them for parts and if something missing then only that part is made new, not everything and all. The same thing with the crystal sync motors I make: if I can reuse some parts of some older design I will not make everything again and again, just take the old one and modify it. That is why I have been able to make something like 150 designs over the 5 or 6 years I have made these. If making everything out of scratch over and over and over again I would probably have like 4 or 5 designs total. That would also have made them so expensive to design that I would probably only had money to make maybe 2 designs total. That is the difference between reinventing the wheel over and over again VS. inventing different "building blocks" and then assembling different style of devices from them only fine tuning the parts to fit. For example I made the 4-speed crystal board design for my Universal Motors but can use the same board for the affordable NPR motor I am making to save 50% of the costs and reduce developing time from 1 year to 1 month. 

Theoretical research is important too. Of course it is, and there may be some really clever stuff found which makes things easier in the far future. That may not be the thing saving film shooting/originating right this moment however. If wanting people to continue shooting film we need stuff right now and if the fully electronical movement would work fine 4 years from now then it can make sense to start developing it. But don't rely on that as your only way to shoot film. Get something which you can use now. You can make like 30 short films with the existing already working cameras when waiting for the theoretical not-working-well-yet camera to be further researched and hopefully finished at some point.

I mean, it is like post-apocalyptic world. Who the heck would dig out iron ore and start making steel out of that when there is bazillion tonnes of high quality steel from car wrecks and rebar from destroyed buildings etc. around? I am scratching my head here trying to figure out why the guys wanting to make a knife and some spearheads and arrow heads for hunting start building some clay oven and painstakinly processing the iron ore when there is a city full of better quality steel right next to them. Just grab the nearest piece of rebar and heat it and hammer the stuff you need from that 😅  

[Friedemann Wachsmuth]

6 hours ago, Tyler Purcell said:

The math I've done says, a $15k body with $5k magazines is possible, but what owner is willing to risk $30k on a "new" companies camera, even it were really nice?

Where do the remaining 10k came from? Sales margin? 🤔

[Brandon Paterno]

In regards to making cameras from scratch versus buying parts — yeah !! it would be great to simply find cameras with certain harvestable contact surfaces like rollers, pressure plates, hell movement too…

but from my limited experience it seems to come down to individual drive. 

i got a 3D printer for a tenner whereas a K1, BH filmo, keystone, etc cost at minimum a hundred… and then if those cameras work why bother reengineering them? it is fun to mess around with possibilities. and for some of my peers myself included cost is a significant barrier. i just graduated so in between applying for work i’ve been making my situation work for me… community labs exist within two hours and some gas money, and so far i have had substantial difficulty with finding out if these groups cater to more than just intro courses and cost prohibitive rentals. they’re busy! we’re all busy! i got impatient so i sunk a small bit of time into building a very reliable optical printer and have been working towards a dicking-around space to continue working with the film medium with eyes towards specific projection phenomena. like i had the time to double down for a few days and now i have a set of designs and measurements for the adaptation of sixteen millimeter kodak magazines to single perf operation… for a specific section of my film.

i’m not gonna wait for someone else to figure out an answer to my problem, and i can’t afford to outsource and expedite these questions so i make scraps and bobs work. I gotta do what i need to for my film to work the way i want it to!

i hesitate to publish detailed info on the mags because of the diy approach… it isn’t perfect. this camera henry’s built isn’t perfect. it’s a start. i learned a long time ago that printing is just always going to be a small component… the rest will be R+D. henry should publish the design files for his camera, and i should publish the notes i’ve got on the mags. In the wrong hands where someone jumps into film mechanics thinking it’s as trivial as hammering a nail into wood will eff up their build because they don’t understand how complex a film movement is. a lot of people who think DV is analog because it uses a linear playback schema will come to this … people who haven’t loaded a motion picture film camera before. all of these camera movements can be simplified, but there are huge problems of mechanical reliability that these things which are based around extrusion style printing can’t overcome. and for some of us this can be solved with better access to tools and personnel , and by proxy funding. and film— lots of it.

but there’s only a few handfuls of people around + alive globally who have been closely addressing these problems over a career. information for newbies like us is hard to come by. and there’s all sorts of different interests in what to do with the medium. and if the information was out there, how does it get simplified without being oversimplified? really it also seems that to anyone looking to step in to the field that the barrier to entry drive wise ends up being a deterrent even if they propose questions deemed as generally impossible. all this to get to say — which has already been made clear throughout the thread anyways — is that the crushing limiter of new technical approaches to film is time and money, isn’t it?