NAC 16mm Analysis Projector

[carllooper]

Have just taken possession of an NAC 16mm analysis projector. Ex NASA. Here is it's brochure:

http://www.nacinc.com/datasheets/archive/DF-16C.pdf

It's in great condition. No dust. Can't wait to start doing experiments with it. It's a 120V/10A one, so have to wait for a step down transformer to use it here in Australia. Had to refresh my memory on electrical theory: Watts = Volts x Amps. So 120V x 10A = 1200W. So that was what I had to get: a 1200W 240V to 120 volt step down transformer. Found one. It's on it's way.

It has a great open plan architecture. Everything is easily accessible. Lots of open space and simple design. Will be easy to modify for what I want it do - UltraPan8 film to digital transfers.

I contacted the original distributor to see if I could get hold of tech specs for the remote. They said all of the specs are in an archive in Japan and that they couldn't access such. Shame. So I'll have to do a bit of reverse engineering on the remote control if I'm to have the projector computer controlled.

Carl

[richard p. t.]

Hi Carl,
this is too good a projector to use for a telecine unit. With a frame-by-frame telecine projector, which is what I assume you are interested in, you don't want a shutter anyway. The beauty of this projector is that the shutter keeps running at 24 revolutions per second regardless of what fps rate you run it at. So there will always be the flicker problem. I would LOVE to get one of these projectors for performance. Sell it to me instead. Get yourself any old projector and slow it down using a variac transformer. Cut the shutter out and there you are. Please don't chop this one up!
rt

[carllooper]

Hi Richard,

it wasn't just for telecine work I acquired the projector, but for art installation work as well, and just how cool it looks too ! The ability to computer control the projector fascinates me. I'd like to interface it with a Kinect device, making it possible to operate the projector through hand gestures.

Last year I saw Last Year at Marienbad, which I hadn't seen since my youth. I want to make a short work that explores similar ideas but which would run on an 'interactive' projector.

There is this zone, between the world of cinema on the one hand, and interactive media (eg. computer games) on the other, that interests me.

Carl

ps. the mods I have in mind won't be irreversible changes to the projector. They'll be additional capabilities that leave the original system intact. For example, the computer interface I have in mind will plug into the original port respecting the original design. Nothing inside the projector will be altered other than attaching some additional sensors on the shutter (which can be removed). For the telecine work the shutter blades can be regarded as a problem. I wasn't intending to remove them, but if I do I'll do it in such a way that they can be put back. My intention is to actually syncronise exposure with the blade motion, so that the blades have no effect on exposure - thus the need for a blade sensor - which will basically be an LED light source shining on a light sensor, the flow of which is interrupted by the blade, producing an on/off signal that can be tracked by software.

[Nicholas Kovats]

Your the NAC UP8 man!

[carllooper]

At first it looked like there was a simple selector, on the bottom of the projector, to change it over to 240 volts, but on closer examination of the internals I noticed that only one of the leads from the selector was connected to the remainder of the machine. The internal transformer was for 120V usage only. So it wouldn't be just a case of rotating the voltage selector and plugging it in. I'd either have to change the internal transformer (and god knows what else) or add an additional setp down transformer.I decided the simplest solution was to leave the machine as is and use an additional external step down transformer. As far as the machine was concerned it would be running as if in the US. The only difference is that the power supply would be running at 50 Hz, rather than 60Hz, ie. the speed of the machine control be a little different from normal operation. But given that the machine is designed for variable speed through a remote control that meant there wouldn't really be any problem. All I'd have to do is adjust the speed differently from the markings on the remote.

Reverse engineering the remote control is going to be a chore. I think I'll leave that for a while. I can just set the projector to run at 1 frame per second (or slower) using the existant remote control and simply slave the camera off the projector.

As mentioned in the previous post I'll be attaching a sensor to monitor the shutter blades. From what I can tell the claw operates in sync with the shutter blades. If I rotate the blades by hand the claw recipricates at the same rate, so I assume it's not the case that the blades are spinning independantly of the claw motion. In other words if the film is travelling past the gate at one frame per second the blades would also be rotating at a reciprically slower speed. unless there is some disengagement mechanism I can't see.

Interestingly there is a separate motor driving the film feed. Looks like there is digital circuitry used to regulate syncronsiation between the feed motor and shutter/claw motor. Interesting. I wonder why they've done that? I expect it's just additional rigour - two motors being better than one. But then there's more opportunity for calamity - for example, were the digital control system to fail. Hmmm.

The architecture of my camera rig, with bellows and enlarger lens, interfaces well with the projector. Nothing getting in the way requiring a mirror or hacksaw to solve.

The lamp is easy to remove and replace with LED lighting. What is so cool is that it's built in a really functional way rather than being stuffed into some designer's idea of what it should look like. Just simple no frills rectangular casing. And inside is even more beautiful. Everything is where it should be for function rather than space-saving. I nice workhorse of a machine.

The gate aperture is wider than a normal film projector, which is great. Indeed the gate aperture makes the sprocket hole entirely visible. How cool is that. The claw is a double claw - with one claw engaging a sprocket above the currently displaying frame and one below.

There is a frame adjust knob on the top of the projector for moving the claw up and down relative to the feed. This could come in useful for multiple exposure of a frame in different positions. I could put a stepper motor on the claw adjust and jitter it up and down relative to the camera, oversampling the frame to compensate for the bayer filter on the camera.

Carl

[carllooper]

The brochure says "flicker free at all projection speeds" but I can't quite see yet (until I get the step down transformer) how that would work. The blades would have to spin independantly of the claw drive.

According to the brochure it can hold a single frame indefinitely, and advance a frame at a time. Perhaps when it's doing this the blades don't spin at all. Anyway - will see in the next few days how it does in fact operate. Spinning or not the blade sensor will do the trick.

The camera I'm using is a Canon EOS under custom software control that I've written using the Canon EOS SDK. The software can operate on the entire range of EOS cameras so when I've got the dollars I can upgrade the camera to a higher definition EOS. My current camera, a 1000D, can do RAW 4K capture at a maximum of 3 fps, but I'll be doing long exposures to start with (using low light levels). And multi-sampling to get around the bayer filter limitations and increase the dynamic range of the result.

I've programmed the EOS so that the mechanical mirror just stays open indefinitely, ie. it's not opening and closing per frame capture. The mode is called "Live View". It's normally used to facilitate an SD video signal going to the camera's video display. But I'm just using it to avoid stress on the mirror. Otherwise you eventually kill the mirror, as others have experienced testing or using such cameras for datacine work.

From there the data will go through the super-resolution processing and other strange techniques I've been developing.

Carl

[carllooper]

Taking cues from VideoFred and friends am looking into LED light sources for the illumination

Found some RGB LEDs to use as a light source here:

http://www.sparkfun.com/products/105

These are much better than white LEDs as you can create a much wider spectrum of light and control the colour to exactly what you want.

Found a board into which the LEDS can be plugged:

http://arduino.cc/en/

And programmed:

http://www.arduino.cc/en/Tutorial/ColorMixer

[Bjorn Andersson]

As Richard earlier have told, the shutter will always run in an constant speed of 24f/sec. The claw are stopped from entering into film by a solenoid when runing on lower frame rate than 24 and only "release" when need. This solenoid can sometime be "dirty" and not move fast enough. A cleaning of the "sticky" solenoid solve the problem.

[carllooper]

As Richard earlier have told, the shutter will always run in an constant speed of 24f/sec. The claw are stopped from entering into film by a solenoid when runing on lower frame rate than 24 and only "release" when need. This solenoid can sometime be "dirty" and not move fast enough. A cleaning of the "sticky" solenoid solve the problem.

Thanks Bjorn. So the shutter sensor I have in mind will do the trick. Or otherwise a long exposure of a second or two to cancel out any inconsistencies in the shutter blades.

Carl

[granfer]

I also have one of these machines, which does not respond to the speed control on the remote. ANY technical information on it would be welcome; ANYBODY, please Email or PM if you have any knowledge.
Happy to compile and share all information and experience gleaned with others.
granfer

[Bjorn Andersson]

As I told, the claw relese solenoid use sometime be "sticky" and will not work. Clean it. It use to be some oil/grease inside it which stop the movement. It must be dry.

[carllooper]

I also have one of these machines, which does not respond to the speed control on the remote. ANY technical information on it would be welcome; ANYBODY, please Email or PM if you have any knowledge.
Happy to compile and share all information and experience gleaned with others.
granfer

If and when I get to reverse engineering the remote control I'll certainly post the results. It could be quite a difficult job. Don't know. There could very well be complex signals going across those wires. Will see. My hope is that I can just log the signals going across the wires, cross referenced with what buttons are being pushed, slider positions, and what the projector is doing and then just build a system that reproduces those same signals but through a computer interface. But could very well become more complicated than that. There may be quite intimate feedback signals from the projector going into the remote, that are then processed by the remote in a particular way, before being sent back to the projector. In fact I'm almost sure of it.

Carl

[carllooper]

I would LOVE to get one of these projectors for performance. Sell it to me instead.

Hi Richard,

I'll be getting the transformer today or tomorrow and will give the machine a test - cleaning the solenoid if required. If you'd like to borrow the projector for something I'm happy to come to some arrangement. I'd love to attend anything you're doing. Let me know. A few years back (about a decade ago now that I think about it) I was going out to Daylesford with a friend almost every weekend. My friend was renovating a house out there. A beautiful place. Some eccentric (and slightly scary) characters out there.

cheers
Carl

[carllooper]

As Richard earlier have told, the shutter will always run in an constant speed of 24f/sec. The claw are stopped from entering into film by a solenoid when runing on lower frame rate than 24 and only "release" when need. This solenoid can sometime be "dirty" and not move fast enough. A cleaning of the "sticky" solenoid solve the problem.

I've acquired the transformer and done the first test. The solenoid is clicking away perfectly varying in rate with respect to the speed control - but it looks like the small plate against which it is tapping has snapped off, which is probably not surprising after decades of being tapped. Without the solenoid engaging that plate the projector just runs at full speed no matter what position the speed slider sits. The forward and reverse control works, as does the single frame control (other than not engaging the aforementioned plate) So other than the problem just described everything else is working perfectly. Just need to repair that plate and all should then work.

Addendum.

On further inspection it's not the end of the plate that has snapped off. It's that the plate has actually shifted a little. Or seems that way. The plate controls whether the claw engages the film or not. As it stands the claw doesn't appear to ever engage the film. Looks like there are some factory calibration bolts I can fine adjust to bring the plate back into alignment with the solenoid and the claw back into alignment with the film path.

Addendum 2

Found the culprit. A missing ball bearing in the plate holding the claw. I think the solenoid only has to nudge the claw plate for it to work - which it is in fact doing. But the claw plate needs to pivot on the ball bearing, which is missing - which is probably the real reason why the claw never engages the film. Looks like someone might have tried to put some glue in there instead, either that or the lubricant for the ball bearing had dried solid. Fortunately the claw plate has two ball bearing mounts, one of which is still there. So I can take the existing one in somewhere and get a new sister for it.

Addendum 3

I found the ball bearing! The real problem was that the lubricant had dried up - making the plate not pivot properly. When I loosened the plate, that loosened the dried up lubricant. But I inadvertently allowed the ball bearing to escape in the process. Have put the bearing back in and it's now working. Will need to put some fresh lubricant in there.

All fixed.

Addendum 4

Found that, as Bjorn warned, the solenoid can stick. Unless the end of the solenoid is lubricated, when it pushes on the plate it can become stuck there - the plate actually holding it from returning. As a consequence the claws then remain disengaged from the film. So a little lubricant on that should do the trick.

Carl

[carllooper]

The projector's running well. The way it works is quite interesting.

The shutter blades are always rotating at a constant rate of 24 fps, as is the claw to which it is permanently fixed (it currently runs a little slower than that on my setup as I'm using 50Hz power rather than 60Hz, but there is a second belt gearing wheel inside to remedy that). However the speed control on the remote operates a solenoid which can move the claw into and out of the path of the film at varying frequency, ie. while the claw itself is always operating at a constant rate (the nominal 24 fps), the rate at which it actually engages the film is varied by the rate of the solenoid.

The reason for a separate film feed motor becomes clear. The feed motor is controlled by a pressure sensor near where the film enters the gate. If you push on this sensor it turns off the film feed motor. When you release the sensor, the motor restarts again. The idea is a feedback system. The film is fed at a fixed rate from the feed reel, but where the film enters the gate it is of course, held stationary every nth of a second. The result of this is that the amount of film, between the feed spindle and the gate would increase at an unknown rate as far as the feed motor is concerned. But as it increases it puts pressure on the pressure sensor, which then stops the feed motor. When the claw advances the film by a frame this has the opposite effect, pulling the film off the pressure sensor, which restarts the feed motor. As a result, the length of film between the feed motor and the claw is maintained as a constant (+/- a few mm), no matter what fps the film is playing.

Particularly fascinating is the single frame advance. The claw is disengaged indefinitely. When you push the frame advance button, the solenoid jumps out, causing the claw to engage the film for an instant, moving it one frame, and then disengages just as immediately by the solenoid jumping back in. The precise timing on the solenoid is no doubt a function of the projector's circuitry. Plenty of ICs in there. The dawning of the digital age.

For computer control of the projector a simple idea will be to just create a mechanical interface that fits over the remote control. The mechanical interface (computer controlled) would operate the slider and the push buttons. No need to reverse engineer the remote control pins. Not very elegant but far simpler to implement. If I get a chance to visit Japan perhaps I could dig up the tech specs buried in the archive there.

Carl