Homemade Camera

[Jack Linder]

After pricing the rental rates for a standard panavision, I decided I may as well make my own camera. Many would cringe at this notion but they would of course be underestimating my welding skills. Likewise it would only be somewhat challenging in creating a consisnet timed crystal motor. The same technology that governs my watch, applies to the motor (sort of). Proper calibration and measurment is intergral to this project. Problems do arise however with the proper calibration of the optics. While I will not be making an actual lens, the camera must be able to accomadate one accordingly. I am worried about the flange focal depth and ground glass setting, as if they are offset only one thousandth of an inch the image would be in a state of ruin. Is there anytilng else I should watch out for?

 

I work with SLR cameras. I dont see why one would have to do camera maintenence with a motion picture camera and not a still camera. Does the flange focal depth and Ground glass settings change so easily with a MP camera?

Edited January 16, 2005 by Jack Linder

[GeorgeSelinsky]

After pricing the rental rates for a standard panavision, I decided I may as well make my own camera. Many would cringe at this notion but they would of course be underestimating my welding skills. Likewise it would only be somewhat challenging in creating a consisnet timed crystal motor.

<{POST_SNAPBACK}>

 

The time and energy you spend making your own cine camera will not make this a worthwhile project. It's like saying "I can't afford a Cadillac, so I'll make my own".

 

You can buy a sync 35mm motion picture camera for under $10,000. You can buy a non-sync 35mm motion picture camera for under $5,000 (even under $1000). Making your own that will be on any level of quality will cost more than that in sweat, blood, and materials. You will drive yourself crazy and in the end you'll realize it would have been easier to save up your money and buy one. If anything, you might want to look into getting a damaged camera on Ebay or elsewhere and trying to ressurect it. That might be a worthwhile enterprise if you have the skills and equipment to do so.

 

- G.

Edited January 16, 2005 by GeorgeSelinsky

[Chris Keth]

The time and energy you spend making your own cine camera will not make this a worthwhile project. It's like saying "I can't afford a Cadillac, so I'll make my own".

 

You can buy a sync 35mm motion picture camera for under $10,000. You can buy a non-sync 35mm motion picture camera for under $5,000 (even under $1000). Making your own that will be on any level of quality will cost more than that in sweat, blood, and materials. You will drive yourself crazy and in the end you'll realize it would have been easier to save up your money and buy one. If anything, you might want to look into getting a damaged camera on Ebay or elsewhere and trying to ressurect it. That might be a worthwhile enterprise if you have the skills and equipment to do so.

 

- G.

<{POST_SNAPBACK}>

 

 

I think George is onto the right track. If you're interested in this, try to find a damaged camera on ebay. Repairing a camera is much more reasonable than purchasing one. There's a pretty good reason that a new 35mm camera costs tens or hundreds of thousands of dollars.

[Rob Belics]

Back in the early 80s and on I built my own computers using wire and individual chips. There was no hurry but tremendous pride in creating a machine no one else had. And it didn't cost me much either so I see where he could be coming from.

 

Somebody had to make the first Arri or Panavision prototype and I don't see why an individual couldn't either if he took the time. Plenty of people build their own cars from scratch, too.

Edited January 16, 2005 by Rob Belics

[GeorgeSelinsky]

Somebody had to make the first Arri or Panavision prototype and I don't see why an individual couldn't either if he took the time.  Plenty of people build their own cars from scratch, too.

<{POST_SNAPBACK}>

 

Well, there are teams of engineers that work on these things. They have thousands of dollars worth of metal machining gear and the experience to use it well. They also know about things like metal properties, lubricants, and so on. They make more progress within a given timeframe. People with no experience in this or experience in one area but not another are going to have to go through a lot of trial and error (and expense) to be playing on the same level as the innovators of the Arri and Panavision.

 

If you want to do it as a labor of love, by all means go ahead. I do think it's a good idea, however, to start by repair anyway, because you learn a lot about camera design. If you got good enough you could then start modifying a camera to suit your specific needs. You could for instance use the movement of an Eyemo or other camera, and build a different shell around it, etc.

 

If your main motivation is to make your first feature on a budget, making your own camera from scratch makes as much sense as making your own filmstock.

 

- G.

[David Mullen ASC]

I would at least consider buying the movement and building the camera around that -- we're talking about precision mechanical engineering here, like crossing a watch with a sewing machine. A 35mm frame is magnified many times over so the slightest misregistration, jitter, etc. is quite noticeable.

[Nate Downes]

I've played with the idea of making my own camera as well from time to time, buying a movement and building the rest of the system.

[Adam Frisch FSF]

I had this idea for years and years, even did designs and went to machinists. My design was to be the smallest, lightest carbon fibre camera with detachable mags. It later evolved into a pure 2-perf design because I wanted to minimize the size even further by just having 200ft mags. Much like an A-Minima but for 35mm.

 

Plausible and doable if you had a little more money than I had. But what really proved to be impossible for me - and believe me I searched the world - was finding electronic engineers that could do a digital crystal control unit to drive the motor. It sounds simple, and basic PWM circuits are readily available, but when integratede into a battery operated design it all became terribly hard to find people. I must have asked and sent at least a 100 mails to electronic engineers, none could help me. I could of course have asked specialists like Tobin Cinema Systems to make me some motors, but then I'd have to pay their mark-up which was an impossibility if I ever wanted to manufacture the camera.

 

You must understand, electronic engineers are so specialized that even if they have experience with similar designs, most of the time they can't help you. For instance, a normal portable CD player has basically the crystal controlled motor as to make the disc spin at a constant, measured speed. So naturally I though someone who had had experince with that would be the ideal solution. Wrong. Turns out they don't have a clue as how to CHANGE the speed - that's a completely different field.

 

So I gave up.

[Rob Belics]

I must have asked and sent at least a 100 mails to electronic engineers, none could help me.

You should have sent 101 cause you didn't ask me! :P

 

I don't understand why those guys couldn't help. I probably could have but motors are not my thing even though I once designed vending machines.

 

Go figure. Worked in television, hired by Pixar, worked on vending machines.

[Jack Linder]

Adam, couldn't you have just bought a motor and integrated it into the camera? The camera body and the pull down mechanism would have had to accomadate one, so you would have had to make some compromise in the design. I am approaching this project from a fabrication/welding angle not electronics. I never thought a motor could be so cumbersome.

Edited January 16, 2005 by Jack Linder

[Marty Hamrick]

I have a friend who built some replicas of old hand crank 35mm MP film cameras for display.I forget which model it was (Moy and something ??It didn't ring a bell with me)but it wasn't something that could actually shoot film.It made a very convincing prop though.It took my friend several months to build.She said a functioning model would take over a year.

 

Marty

[Adam Frisch FSF]

Sorry Rob, I wish you'd been here back in those days:-)

 

The motor is the easy part - that's not a problem. The tricky part is to control the speed of the motor - designing the "brain" of the camera. If it doesn't run synch, then it's useless and wouldn't probably sell at all. That's the key obstacle to overcome. You also need to program a user interface like a small little LCD and buttons that shows the speed, voltage, and any other thing you need presented.

 

Not only that - since it's battery powered you want the motor to run at the exact same speed EVEN if the voltage of the battery gets low and even if the load in the mag fluctuates (some magasines run smoother than others, some rolls are bigger and have more friction, wobble etc). That sounds easy enough, but it isn't. The last thing you want is a motor that starts to loose speed as the juice lessens - that's why you need precise circuitry to control all this. Ideally, you just want the motor to shut off below a certain voltage and not run at all (like they do on modern cameras), but then you get into voltage under load problems and other headaches (i.e. the battery showing that it has the juice volt-wise, but when you load it, or run it, the voltage dips below the set figure). Even with the most sophisticated new cameras, you can sometimes encounter speed troubles on very low battery voltages. It happened to me just a couple of years ago.

 

I'm not trying to discourage you, you just need to attack the problem from the right angle. All the mechanics of a camera is easy - any machinist could do that. All the optics are fairly easy, too. It's the electronics that need to be adressed.

Edited January 16, 2005 by AdamFrisch

[GeorgeSelinsky]

The tricky part is to control the speed of the motor - designing the "brain" of the camera.

<{POST_SNAPBACK}>

 

Crystal syncing a motor is really no big electronic task.

 

The motor produces a pulse based on its rpm, my Arri II for instance has a tachometer. You can create an electronic version of this. One way of doing this is to have a disk attached to a motor gear with holes symetrically punched into it. You then use an LED and a photo circuit opposite each other, between the disk like so:

 

LED -> || --> PHOTO RECEPTOR

 

 

When the hole turns up, the LED and photo receptor see each other, thereby producing a current. When it leaves, they don't see each other anymore, hence no current. What you then get is a pulse wave directly related to how the motor is spinning.

 

You then take this pulse wave, and compare it to the pulse of a crystal circuit (you may need to divide or multiply the frequency in order to do a proper comparison, there are IC's for this). In an analog world you use an integrated circuit known as a PLL (Phase Lock Loop). When there is a difference between the motor RPM and the crystal, the PLL puts out a voltage. That voltage can be used to regulate the voltage of the motor, applying maximal resistance when there is a lot of voltage, and when it drops to progressively lessen the resistance.

 

With digital technology, you use an A/D converter and do the PLL operation digitally with gates or a dedicated IC. It's not, from the perspective of an electronics engineer, a particularly tough task.

 

The voltage from the PLL can also be used to regulate a sync alarm light which turns on when the camera is running off speed (or a sync alarm beep, or both).

 

- G.

Edited January 16, 2005 by GeorgeSelinsky

[Guest Sean McVeigh]

I must have asked and sent at least a 100 mails to electronic engineers, none could help me.

<{POST_SNAPBACK}>

 

I'm building such a controller right now for my 2C. The PWM is too noisy right now (switching 10A over 6 feet of cable -- woops), needs to be flattened out to DC, but the tachometer/footage counter side works well already.

 

And as for the original poster's idea for building a camera... if you have the skills.. do it! Even if you never finish, you'll have learned a helluva lot doing so. And as for the arguments that you're better off spending the money buying or renting.. well, I write off a lot of my spare-time projects as "my time isn't worth anything to me, since no one is paying me 24/7 otherwise".

I could have dropped $1500 on a crystal sync motor, but instead, this has cost me on the order of $100 and a couple dozen hours of sweat. Good learning experience, and makes for good resume material if making films is not your day job.

 

Build your own camera! (then build a film scanner when you're finished.. I'll just copy your design ;) )

[Nate Downes]

George, I actually did just that, but instead of using "a disk" I used the shutter itself. I modified the shutter so that it had holes punched on an inner-ring, inside of the frame-gate area. Then I put the LED system on the opposite side of where the film gate was, and viola, one less part needed in my machine.

[Rob Belics]

It's the electronics that need to be adressed.

Without knowing the exact details, for me it would be the easy part.

 

Better accuracy could be obtained using contrasting slits, like barcode, and scanning it with the led but I don't know that would be necessary.

[Adam Frisch FSF]

Most modern film cameras have brushless motors. These are motors where the inversion of the field is accomplished electronically, therefore they don't need to be "monitored" with such a disc device. They run at the speed you "tell" them and if they don't they warn you about it. They're also much more silent and much smaller. But of course, they cost more.

 

Worth noticing is that all the new Arri's are in fact motion control-controlled (?). The shutter is run by a separate small motor completely independently from the main motor. This saves weight and doesn't require cumbersome mechanical linking that decreases efficency. But the control has to be very good - the shutter and camera movement have to move exactly in sync or you get into trouble. This requires a computer. In fact, on the 435 you can turn the mirror away in a service menu and not have it move at all but still run the camera. This creates some very interesting streaking/smearing effect and was the rage amongst some of my fellow music video DP's some time ago. Problem is you don't see s*** in the eyepiece!! :D

[Guest Sean McVeigh]

George, I actually did just that, but instead of using "a disk" I used the shutter itself.  I modified the shutter so that it had holes punched on an inner-ring, inside of the frame-gate area.  Then I put the LED system on the opposite side of where the film gate was, and viola, one less part needed in my machine.

<{POST_SNAPBACK}>

 

Just to add 2 more cents to this thread, I interfaced the tachometer at the inching knob. 1:1 revolution-to-exposure ratio there. On a 2C there are actually a couple of mounting holes right around that area, so it's easy to install/remove.

[Alessandro Machi]

If you take your concept of making a 35mm camera as small and lightweight as possible that does single frame, time-exposure, and intervalometer, (real time filming not required), I would be very interested in purchasing one.

 

Emulate a Eumig 881PMA in 35mm, that would be awesome.

[DanStewart]

It's also worth noting that in the DI world it is feasible to build a camera with terrible registration if it saves on cost.

 

I have some work done on a servomotor driven design that could do high speed and change frame size electronically (2- 4 perf, switching out the gate of course). It's registration is awful but it puts a sync dot between the perfs aatoncode style so as long as you're scanning the neg you can software register perfectly in post. It would also be extremely quiet since there is no mechanical movement.

[Rob Belics]

But you've got to pay for that software use don't you?

[Guest Sean McVeigh]

But you've got to pay for that software use don't you?

<{POST_SNAPBACK}>

 

not if you also list software designer alongside camera designer as your list of talents ;)

[Nate Downes]

you know, after this discussion, I got curious about finding the film gate mechanism for my favorite camera, the Mitchell BNC. You know, I can't find one for sale anywhere.

[Guest Sean McVeigh]

you know, after this discussion, I got curious about finding the film gate mechanism for my favorite camera, the Mitchell BNC.  You know, I can't find one for sale anywhere.

<{POST_SNAPBACK}>

 

Probably tough to find without making a lot of phone calls.

I have seen movements for sale on ebay in the past.. they don't keep a long enough history to allow me to find them though...

Last week there was a Mitchell G.C. Hi Speed camera movement (1-120fps) including gate and pressure plate for $600 (auction listing is here) apparently it didn't sell.

 

Stuff is out there it seems.

[Nate Downes]

What is the difference between the BNC, GC and S35R anyways? I've looked for info, but never run across any answers.

Edited January 20, 2005 by downix