While artists all over the world continue to work with celluloid film, they are often confronted with precarious screening conditions due to increasingly old and hard to repair equipment. In particular, film projectors and their aging mechanical parts have become less dependable, contributing in many cases to the destruction of the film material instead of ensuring its optimal presentation. The last 16mm commercially available film projector was built in the 90’s but most artists, archivists and projectionists have to deal with much older equipment, sometimes going back to the 60’s and 50’s. At the same time, the traditional industrial manufacturers have disappeared or shifted to other fields, service personnel have retired and spare parts are rare and unreasonably priced. Designed mainly for standard film projections, vintage projectors also fail to cater to contemporary demands: While not offering enough flexibility for artists who work in expanded cinema, they also don’t usually meet the needs of archival projection. Along with the digital revolution, which has, in the last decade, already greatly reduced the opportunities for experiencing analogue film projection, the aging technology of vintage projection equipment has become an important factor in the disappearance of analogue film projection and the uniqueness of its experience.

Our idea is to develop a state-of-the-art and modular 16mm film projector using only open-source technologies and non-proprietary/commonly available spare parts. We believe that especially the central mechanical elements of the old projectors – claw mechanism, shutter wheel and film transport - are in most cases so well engineered that a new development here would be a waste of time and energy. Instead, we want to build the projector on the basis of an existing - and easily available - projector mechanism. The same applies to the optics: lenses that are compatible with projectors made by Eiki, Bauer, Bell & Howell and Hokushin are available worldwide in good condition. This projector should cater to the needs of contemporary film artists, archivists and projectionists alike.

Design

• Modular design
• Open source technologies
• Non-proprietary & commonly available spare parts (3D-printable)
• Adjustable height & tilt
• Lightweight for travelling & portability during projection
• Option to project in vertical format (90° tilted or Prism system)

Power

• 110V & 220V
• Optional: Battery for outdoor & portability during projection

Light source

• Super bright, dimmable LED
• Color temperature adjustment for differently timed prints,Tungsten or Xenon or redshifted film prints
• Digital shutter (flicker)

Film formats

• 16mm – Super-16 – Ultra-16 – open gate
  (with switchable format masks)
• Steady focus between print & reversal stocks
• Optional: interchangeable sprocket wheels for shrunk film

Optics

• Wide zoom range lens 25mm – 150mm
• Compatible with Bauer, Eiki & B&H lenses (adapter tubes)
• Focus with Worm Gear
• Anamorphic lens holder
• Holder for Elmo Viewer Type 100 (viewing without screen)

Transport

• Crystal sync speeds: 12 – 15 – 16,66 – 18 – 23,976 – 24 – 25 – 29,97 – 30 FPS
• Manual vario-speed from < 1 to 30 FPS
• Vario shutter wheel independent from FPS
• Digital frame counter
• Memory counter for in- and out-point
• Fast rewind in both directions

Audio

• Optical & magnetic audio (no built-in amplifier – just outputs)
• Microphone input for live voice
• Headphone jack
• Integrated digital audio sync system

Connectivity

• Sync with digital audio, video & midi
• Sync between several projectors
• Switchable from master to slave
• Sync with Elmo ESS system
• Optional: ready for telecine
• Remote control: IR / Cable / Bluetooth

Accessories

• Development of a compatible looper device
• Spool arm extensions

Our project takes place over a period of two and a half years, and should be completed by September 2025 with the presentation of a prototype at the Back To The Future Festival in Rotterdam.

As a first step, in a team of two, we disassembled four film projector models, which we found offered a suitable mechanical system that could serve as the basis for further development. We have defined three fields of development, for which we will have to collaborate with different experts. These fields are light source, film transport mechanism and electronics.

But before we can take the next step, we also realised that we are at a junction where we first have to decide which path to take:

A. Develop a flexible upgrading system which suits various existing projector models. This would ensure that artists would be able to upgrade their own projector model, no matter where they live and what projector they own. Our concern is that it may be difficult to develop parts that can adapt with various existing mechanical parts.

B. Develop an upgrading system for only one widely available projector model. This would enable us to develop much more specific parts and create an integrated concept for that one type. The downside is that many projector models are not equally available in all parts of the world.

C. Develop a DIY-Kit, replicating mechanics from various existing models, using techniques like 3D-printing , CNC- and laser cutting. This would enable artists all over the world to build up their own, modular and state-of-the-art projector from scratch. Eventually we could provide and ship readymade parts that would be too difficult to produce individually.

Once this crucial decision has been made, we want to bring an expert in electromechanics on board to accompany the project through to the prototype. At the same time, we want to build an online community with whom we can share our ideas and who can also test and improve individual parts. We are already in contact with several people who are independently working on similar developments and who are waiting to finally share their knowledge and experiences with each other. In the end, we will hire an industrial designer to work with us in order to build a prototype.

In the following we show detail shots of various 16mm projector models that we have disassembled for better examination. For each model we list the advantages and disadvantages that we noticed during disassembly.

1. Siemens 2000

2. Kodak Pageant

3. Hokushin SC-10

4. nac Analysis Projector

5. Eiki RT2

In August 2023 we decided that before building a prototype, we first needed to find an LED light source that is capable of replacing the common 24V 250W halogen bulb. Chzech film artist Jan Kulka recommended high density LEDs from Chinese manufacturer Getian, which he has used for his Archeoscope expanded cinema machine. Over a period of 6 months we tested several LEDs with different Wattage, starting from 200W, then moving up to 400W, 600W and finally 800W. We based our decisions not only on the brightness but also on the size of the LED chips in relation to the 16mm projector gate.

In order to achieve comparable and realistic testing parameters, we took out a 16mm gate with lens holder from an old Bell & Howell projector. We placed the LEDs as close as possible to the gate and step-by-step turned up the voltage in 0.5V increments. We then measured the temperature on the chip, and the brightness of the projection, using a lux meter.

After a few attempts we noticed that we were always reaching a critical temperature of 60°C (recommended by the manufacturer) even before the LEDs reached the rated power capacity and the highest brightness. We realised that we would have to find a more efficient cooling system then the computer processor heat sink with fan that we had started with. The next step to achieve more brightness was to try out water cooling. For our tests we chose a water cooling system (AIO) that is usually used for gaming computers. This actually allowed us to operate the LEDs with much higher voltage and up to their maximum capacities without over heating. Finally with an 800W LED we were able to reach close to twice the brightness of a 24V 250W halogen bulb.

Reference Brightness: Bell & Howell 16mm Projector with 50mm f/1,4 lens at 24 FPS: 10,000 Lux

LED-Setup: Lens: 50mm f/1,4, distance LED-gate: 2,3 cm, distance lens-wall: 155 cm

Test 1 07.08.2023 LED 200W max. 13,6 A 12 - 16 V with air cooling

Test 2 14.10.2023 LED 400W max. 10 A 42 - 48 V with air cooling

Test 3 16.10.2023 LED 800W max. 15 A 45 - 54 V with air cooling

Test 4 29.10.2023 LED 800W max. 15 A 45 - 54 V with AIO water cooling

Test 5 16.12.2023 LED 600W max. 17 A 36 - 42 V with AIO water cooling

Test 5 16.12.2023 LED 400W max. 10 A 42 - 48 V with AIO water cooling

.

After coming to the conclusion that the 800W LED was the most appropriate for our project, we decided that the next step should be to test the LED with a projector that had a working transport mechanism which would allow us to run a test film through the projector and check the quality of the projection. This would mean that we would have to find a projector that we would need to modify in order to fit the LED and cooler. This brought us back to the question of which path to take with the project in general. It became clear that in the time frame that we have for the project, it was most important to be able to test a system where we had variable control over the light and the transport mechanism and that we should focus first on these two parameters. That meant focusing on replacing the lamp and motor in an already existing projector instead of trying to design an entirely new projector from scratch or instead of trying to engineer a hybrid model with the best features of various projectors. So we decided to modify the Eiki RT model because of its availability, robust construction and the amount of space that it offered in the housing for modifications. The simplicity of its transport mechanism was also the factor as well as the idea that the modifications we would make could be easily reproduced by others.

At this stage we decided to bring in somebody with more expertise with projector construction and modification. Based on his experience building his own custom projection devices we asked Jan Kulka from Prague to join our team and continue the development of the prototype with us. In April 2024 we met in Berlin and decided on the next steps to follow. Even though our wish list contains a vast number of possible functions, we decided to focus initially on replacing the motor and installing a flickering LED Lamp that would replace the mechanical shutter.

Jan Kulka took over the technical engineering of our first prototype. The aim was to modify an Eiki RT-2 projector and implement the following features:

• The prototype should be capable of projecting 16mm film at variable frame rates from 0 - 30FPS
• Our 800W high density LED should be installed, cooled by an AiO water cooling system to avoid blowing the LED or burning the film
• The original shutter wheel should be replaced by an entire digital approach through flickering the LED
• The prototype should be ready to be presented at ALUD festival in Barcelona by end of October 2024

At this point of time, we knew that our SPECTRAL project partners from Mire (Nantes) had already appointed experts for their “wandering devices project” to work out a modification of the same projector model, similar to what we had in mind. Their design, created by Zach Poff, Stefan Voglsinger, Guillermo Tellechea & Loic Verdillon can be seen here: Wandering 16mm projector retrofit. Since all four devices projects within the SPECTRAL project are meant to be open source and to be replicated and improved by others, we decided to take over larger parts of their concept as a basis for our prototype. So we discussed with Jan that it would make sense to take over the code, the choice of the encoder, the motor model and the ‘virtual shutter’ system from their wandering devices project. This would give us more time and capacities to focus on the implementation of the 800W high density LED and the AiO water cooling system.

Jan decided to use only basic tools like a power drill, angle grinder and basic hand tools. Only in one case an ultrasonic saw had to be used. In order to make sure that anybody could do the same modification without the need for special machinery, no 3D printing or laser cutting were being used. Due to heat resistance reasons, most custom made parts were made of aluminum. Jan and the technicians he employed did a marvellous job as the following technical details and photos reveal:

1. Eiki RT-2 projector - original parts that have been removed

• mechanical shutter & sill image clutch & dowser
• power supply
• motor
• fan
• all electronics
• sound board
• lower part of the body
• to accommodate the hoses of the water cooler, some holes had to be drilled & grinded into the inner part of the body

2. New light source

• Getian High Density 800W COB LED Chip 55000lm
• Mosfet – semiconductor for switching the 800W LED
• three big resistors for safety & better control of the LED
• extra fan for cooling the resistors & Mosfet
• Condensor lens (probably taken from a Kodak Pageant 16mm projector)
• AIO water cooling system: Fractal Design LUMEN S28 Water Cooling Unit
• yet to come: fan for gate, fan for space in between LED and condensor

3. Two new power supplies

• Meanwell 1000W UHP-1000-48 with built-in current & voltage regulation (only for LED)
• Meanwell 500W UHP-500R-12 (for Motor, Coolers & everything else)

4. New motor

• Quicrun Fusion SE brushless system for Crawler 2-IN-1 with FOC (field oriented control) driver embedded within the motor itself, for accurate positioning & control, big torque at slow & fast speeds, compact size & light weight
• Motor pulley 16 teeth (GT2 type)
• Main shaft pulley 100 teeth (GT2 type)
• Timing belt (GT2 type)
• Magnetic encoder / neodym magnet mounted on main shaft – detects angle of main shaft for precise motor control & frame counting

5. New control electronics

• Main circuit board accommodates:
  • ESP - main controler
  • step down module to 5V
  • inputs & outputs to control panel potentiometers & buttons
  • Relay to switch optical sound LED
  • PWM signal output for the main LED power supply – to control the output voltage/current for the main LED chip (dimming)
  • PWM signal to the motor driver – to control the motor
• New soundboard (simple pre-amp for line out) for optical sound, using Eiki original light sensor & small LED light source (replacing original exciter lamp)
• Arduino Nano for resetting the ESP & Motor driver when Machine gets switched on

• The main LED Chip is switched on & off by Mosfet HY 4306 (mounted on new aluminium heatsink with a small fan), controlled by ESP-Wroom-32 - this serves as the electronic shutter
• 800W LED is powered & dimmed by Meanwell 1000W UHP-1000-48 power supply. The output voltage is controlled by an external VC (voltage control). This VC is generated by our custom made board, which is receiving the PWM signal from the ESP and translates it into 0-10V VC, which from there is being sent to the power supply's own output regulation system (no extra dimmer needed, dimming can be programmed digitally)

6. Buttons & switches

• Eiki main switch functions (from left to right)
  • Reverse Play - Risk mode (playing at low fps may burn the film)
  • Reverse Play - Save mode (LED gets dimmed at low fps, no risk to burn the film)
  • Standby (electronics & fans on) – projector is ready to run 5 secs after plugged in
  • Forward Play - Save mode (LED gets dimmed at low fps, no risk to burn the film)
  • Forward Play - Risk mode (playing at low fps may burn the film)
• Control panel buttons & switches (from left to right)
  • Potentiometer - motor speed (0-28 fps in save mode, 0-40 fps in risk mode)
  • Potentiometer - Ramp for motor speed (variable responsiveness (or delay) of speed poti, from instantaneous-15 secs)
  • Push button (red) - 1 frame backwards when motor is stopped, freeze frame while motor is running
  • Push button (black) - 1 frame forward when motor is stopped, freeze frame while motor is running
  • Potentiometer - LED brightness (off - maximum)
  • Potentiometer - ramp for LED brightness (responsiveness (or delay) of LED poti from instantaneous-15 secs)
  • Potentiometer - number of virtual shutter blades (or flashes per frame) - 1-2-3 blades
  • Potentiometer - virtual shutter angle (duration of each flash) – from closed to completely open
  • Switch (not yet in function)
  • later: 2 potentiometers for sound volume & tone.

water cooling system, 800W LED, copper plate for mounting, condenser lens (in the back)

testing the condenser lens

allocating space for the hoses of the water cooling system

grinding a passage for the hoses

the mounted radiator of the water cooling system

the water cooling head

mounts for LED and condenser lens

all custom made parts for mounting LED and condenser lens

all custom made parts for mounting LED and condenser lens

positioning the copper plate

the copperplate holds cooling head and condenser lens as well

LED and condenser lens in action

the new motor

mounts for the motor

mounts for the magnetic encoder

magnetic encoder

magnetic encoder mounted

magnetic encoder in position

the motor in position

aluminum frame to hold two power supplies

two power supplies mounted

aluminum base with power supplies mounted, heat sink for mosfet & resistors added

aluminum base with power supplies mounted, button panel added

aluminum base, heat sink, aluminum plate to hold electronic boards

button panel with potentiometers and push buttons

.

As part of ¡Alud! #4, a festival organised by our project partner Crater Lab, which is entirely dedicated to expanded cinema, we presented our first prototype at a so-called feedback loop, a public gathering where makers & experts are having a detailed look at the various technical achievements during the Devices project. All four groups showed their latest version of their various prototypes in an exhibition kind of set-up. A few participants who could not attend in person, like Zach Poff and Matt McWilliams in the US, were joining the discussion via streaming. The open and hands-on character of the event allowed a direct exchange of feedbacks about technical features and solutions between all participants, on-site as well as online. Due to our logistic restrictions across countries, this was the first opportunity for our research team to see our prototype in a direct, side-by-side comparison with a similar, unmodified projector model with the original 250W halogen bulb, both projecting the same film. It was amazing to see the brightness and versatility of our projector, nevertheless we also observed a few flaws. This is what we found:

• Brightness: The projection of our prototype was much brighter compared to the 250W halogen projector
• Colours: Even though our 800W LED only has a CRI of 70, the colours looked saturated and vivid. The often debated importance of the CRI value doesn't seem to be as relevant in projection as it definitely is for scanning and colour grading in post production
• Features: All variable transport functionalities worked like a charm
• Optics: Since we are currently using an obscure condenser lens, we should work out a better optical system that can easily be replicated by anyone who wants to modify their own projector
• Mechanics: The projector seems to be very sensitive to tape splices, even though we took extra care to make sure they are accurate. Also Jan pointed out that he had come across one vintage film roll that didn't even get moved by the claw mechanism. Quick adjustments of the claw position did not solve the problem, we will have to do a complete CLA before we can go ahead. For the final release of our projector, we should also consider including sufficient information how to solve such problems
• Projection: Especially in the comparison with the original 250W halogen projector, we found the projected image flickering too much. In direct comparison, the much dimmer image of the halogen projector looks much smoother. This happens especially when the 800W LED is set to higher brightness. But the virtual number of shutter blades and the virtual shutter angle also play a big role in here

We discussed the possible reasons for the flicker:

• Electronics: Zach Poff suggested that there could be a problem with the output from mosfet. We will have to measure with an oscilloscope whether there are any flaws in the electronics
• Sync: It could be a sync problem between the pull of the claw and the flicker of the LED, especially after the prototype had been shipped in a van over a distance of 1700 kms from Prague to Barcelona. The neodym magnet that is attached to the main shaft, only holds by its own magnetic force. So a microscopic shift of its position on the shaft could have caused a slight jitter in the projection
• Human perception: It is possible that with the brightness of our LED, we may have reached a certain threshold, at which the image flicker appears more apparent to the human eye. Matt McWilliams pointed out that back in the days when drive-in theatres in the US increased the brightness of their projections, they also upgraded from two to three shutter blades to compensate for the occurring flicker
• It could be possible that all three reasons may be adding up upon each other and that way cause more flicker

.

Bernd Lützeler filmi@gmx.de

Juan David González Monroy jdgonzalezmonroy@gmail.com

SPECTRAL is co-funded by the Creative Europe program of the European Commission.

SPECTRAL is co-funded by the Kofinanzierungsfonds of Senatsverwaltung für Kultur und Europa, Berlin