Am I crazy?

[Matthew W. Phillips]

I got the bright idea to build my own Super 8 projector from scratch and I want it to be capable of single frame projection for scanning. I have a ridiculous looking functional model that can now project single frames but cant sync and has no shutter. Is this a foolish venture? The images look pretty good bounced off of a black television screen.

 

Anyone else try crazy stuff like this?

[Matthew Buick]

You are not crazy, tis a noble, honourable, and gallant act, I salute you !

[Mike Rizos]

I think it's a good idea as an experiment, although I wouldn't know how to go about it. Could you post a picture of your ptototype?

[Matthew W. Phillips]

Could you post a picture of your ptototype?

 

You want it, you got it...although my digital camera is buried in storage so you have to settle for webcam grabs...so sorry. Here it is.

 

 

 

I will try to get better images in the future, but this is the general idea.

[Scott Bullock]

Crazy? Like a fox, maybe! Super-8 will continue to be a viable format due to the efforts of tireless technicians such as yourself. Excellent work!

[Matthew W. Phillips]

Crazy? Like a fox, maybe! Super-8 will continue to be a viable format due to the efforts of tireless technicians such as yourself. Excellent work!

 

Thank you Scott for the encouragement. However, anyone have an idea of how I can (cheaply) devise a method of syncing for 24fps (or any particular rate for that matter)?

 

No idea is too silly for me to try!

 

I also wish to tell Jim Jannard that I have posted my prototype so I would love to see his for RED. Dont be ashamed of the beast Sir, be proud of it!

Edited September 15, 2006 by Matthew W. Phillips

[Michael Collier]

Its single frame? What advances the projector, a button or a timer? A cheap way to go about it (and also get 4:4:4 uncompressed HD out of the deal) could be to rig a D20 or similar digital camera (still, not video) and a timer to advance the projector. then you would have a couple options. Rig the timer to a delay to snap a photo a few 1/10 of a second after the projector starts projecting and then a second timer to let the system wait for the camera to be ready again.

 

A simpler (but maybe less effective method) could be to rig a photocell/transistor circut to the camera through its external 'shutter' release cable (a simple relay can trip the shutter, and the photocell can trip the relay) The only problem is 1 the photocell would have to be somewhere the light from the projection can reach the photocell (which means it could be in frame, unless it is made sensitive enough to be tripped by bounce light) and 2 if the frame is a darker shot, there may not be enough light to trip the relay.

 

I can help a bit with circut design, if your unfamiliar with electronics. I am not terribly good at design, but I think at least good enough to rig something like that (best of all things like transistors and voltage regulators can be had for free from the manufacturers, if you tell them theres a chance it could be made into a marketable product. I once got a shipment of almost 1000 transistors in different models and several different IC chips for free, including shipment. I made a 2000w amp, though not my design. brilliant design by the EE professor at pheonix university.)

[Matthew W. Phillips]

Its single frame? What advances the projector, a button or a timer? A cheap way to go about it (and also get 4:4:4 uncompressed HD out of the deal) could be to rig a D20 or similar digital camera (still, not video) and a timer to advance the projector. then you would have a couple options. Rig the timer to a delay to snap a photo a few 1/10 of a second after the projector starts projecting and then a second timer to let the system wait for the camera to be ready again.

 

I may aim for this idea...it is more time concuming, but the payoff is worth it, I feel. I am definitely not expecting a marketable product out of this. Maybe something that I can benefit from or at least have fun doing when I drink... :D

[Marty Hamrick]

Its single frame? What advances the projector, a button or a timer? A cheap way to go about it (and also get 4:4:4 uncompressed HD out of the deal) could be to rig a D20 or similar digital camera (still, not video) and a timer to advance the projector. then you would have a couple options. Rig the timer to a delay to snap a photo a few 1/10 of a second after the projector starts projecting and then a second timer to let the system wait for the camera to be ready again.

 

 

You've just basically described a J/K optical printer.

[Michael Collier]

I got sorta interested in this and drew up a quick scematic illustrating a few concepts for control circuts. One that is cheap and simple and effective, and another that allows for a display to show the number of frames or seconds or feet scanned so far, and allows you to set the starting point manually (as well as pause the operation). that one is a bit more complicated, as it requires programing a microchip with a simpl program.

 

Let me know if you want the designs for either of those. The cheap one would have one relay to switch the projector to change frames (and can either be a momentary switch close, or close until a signal is sent from the projector to say it has finnished changing.) and a relay to connect to the cameras shutter release. let me know if you want them because I have designed them, but have yet to do the math to determine resistor and capacitors values for the 555 timers.

 

And it doenst ever have to make it to market, its just a check box you click, and you can get the 555 timers (3 of them, usually $0.55 each, or $0.75 for a chip with 4 built in), transistors, voltage regulator and the flip/flop chip for free. You would have to pay for resistors and capacitors. Assembly would be easy enough (if you know how to solder)

 

The cheap one can probably be built for 25-50 bucks (depending on transformers, relay options, project box etc) less if its just an open perf board. Voltage would be around 5v on the board, 7-9v from the transformer, and easily under 500mW (with relays activated) so risk of electrocution is nearly none (I can't think of a situation where it would be possible, but take this as a disclaimer.)

[Matthew Buick]

You've just basically described a J/K optical printer.

 

Can you get Super 8 Optical Printers ?

[Guest Glenn Brady]

Can you get Super 8 Optical Printers ?

 

Yes, you can. Jaakko Kurhi (www.jkcamera.com) makes an optical printer that can accommodate regular 8mm, Super 8, 16mm, Super 16, and 35mm. I once owned an earlier version of the printer now shown at his website and with it produced reductions from 16mm to Super 8, straight copies, and multiple frames within a frame on 16mm. Jaakko also does conversions of H8 RX and H16 RX cameras to DS8. Great stuff!

[Matthew Buick]

Is an Optical Printer in any way dangerous ?

[Michael Collier]

Is an Optical Printer in any way dangerous ?

 

No, and neither would my drive circut, if its in a box. I was just saying if he wanted to save money the circut could be built without a box, but obviously that means exposed electronics. Of course, the power needed to run the drive circut would be less than it takes to power a flashlight, so I figure if you can't electrocute yourself on a flashlight, you should be fine building my drive circut without a box, so long as the transformer (wallwart) you build it to doesnt supply more than 9vdc at 250-500mW (worst case if built right, imagine what happens when you lick a 9v battery. not much)

 

The drive could be modified to accept a sync pulse from the projector, but you need to make sure the digital camera can take a picture and be ready for the next one within 1/24th of a sec (better if its within 1/48th of a sec) a longer exposure means you can dial down the ASA in the digi camera, meaning cleaner images.

Edited September 16, 2006 by Michael Collier

[James Erd]

A simpler (but maybe less effective method) could be to rig a photocell/transistor circut to the camera through its external 'shutter' release cable (a simple relay can trip the shutter, and the photocell can trip the relay) The only problem is 1 the photocell would have to be somewhere the light from the projection can reach the photocell (which means it could be in frame, unless it is made sensitive enough to be tripped by bounce light) and 2 if the frame is a darker shot, there may not be enough light to trip the relay.

 

I think your right, but I would mount the photocell inside the projector....

 

The photocell could be mounted inside the projector, off axis with the blade of the projector serving to interrupt the light to the cell. The signal could trigger the camera in the same way as photographers use one strobe as the master, and slave the other strobes to it. You might even be able to kobbel some standard off the shelf components to do it.

 

[Light Source]D~~~~~~ [blade]~~~[Photocell]=======signal to camera====I

I I

I I

[Light Source]D~~~~~~ [blade]~~~~~[Film]::::::::(lens)<<<<<<<<<<<[Camera]

 

Another route would be to replace the light source in the projector with a strobe, and have the camera trigger the strobe.... when the strobe flashes it triggers the projector to advance one frame.

 

The difference may seem insignificant at first, but it allows you to build a much simpler apparatus. You should get fewer dropped frames this way. It also eliminates the use of a photocell. That's one less item for you to make or find. Strobes are ubiquitous and easily triggered by cameras. They are better suited for time critical applications which is why Harold Edgerton invented them. They are also not as prone to temperature shifting over time as incandescent. So your results should be more consistent.

 

Also using a photocell for a trigger could be problematic. There are all the issues of what threshold it takes to set it off. Further the use of a strobe gives you the opportunity to simplify the projector even further because the projectors shutter becomes redundant and can therefor be eliminated. This save you the trouble of inadvertent shadowing of the frame by the projectors shutter, due to the vagaries of mixing digital and analog technologies.

 

The only timer necessary would be the one to trigger the camera... assuming that the projector is fairly consistent in the time it takes to advance to the next frame. One more thing to consider is if the camera gets triggered before it is ready it simply does not take a picture, and so the projector doesn't advance, and the situation is corrected on the next cycle.

 

If you want to do it right you would also eliminate as much air as possible. The further the light travels before it gets to the chip the more it is going to scatter. This can only reduce your contrast. So I would recommend getting the sharpest macro lens you can. Remove the projectors lens and focus you digital camera rite on the film. You might even think of this setups as a very low power digital microscope.

 

I've been want ting to do some thing like this with my old Siemenns Dual 16-mm, but I haven't found the right lens or camera yet..... with in my budget that is...

 

I don't think it's crazy at all.

[Michael Collier]

I dropped the idea of a photocell. I thought it would be easier for a non-electronics person to build, which it may be, but it would not be reliable, like you said.

 

I thought of the strobe idea. the problem I see(especially with CCD cameras) is that the rolling shutter can make the timing difficult (even gen locked, look at how hard it is to sync two cameras with a frame sync). The signal comming out of a cheap prosumer camera would need to not only have the blackburst seperated from the image signal, but would need to be delayed to flash the strobe in the middle of a frame integration (CCD talk for expose) 1 or 2 frames later(since processing would make the blackburst come out of camera long after the CCD exposed the frame, it could only be used as an indicator of pacing). accuratley setting exposure and color timing options would be more difficult to control as well.

 

The short flash of a strobe means timing would have to be perfect. For a professional level camera that means a crystal sync timer connected to something that generates blackburst, both of which would be more costly to buy and more difficult to design than my timer. The timer design relys on simple (and cheap) RC (resitor-capacitor) timing, and since its a closed system, small variation in timing (being not crystal synced) would only make the transfer take a few more or less seconds over 100ft role, depending on the ambient tempature (not affecting the playback when brought into the computer, since one film frame will always corrilate to one frame in the camera) It would never drop a frame or lock up.

 

The strobe, since its operating at the same rate as the camera, could lead to problems (since I am guessing his projector is a wild motor design, though theoretically a crystal sync motor may still show problems). The continuous and time-dependent exposure of a video camera means that stop/start control is not possible in the short time it has to move the next frame. A slowdown of the projector by just a little bit would start to drop frames. like your strobe idea, the shutter can be eliminated if desired on this design (the camera would only expose when motor shaft is in the right position and stopped, so seeing the frame move would not affect quality)

 

The timer idea I had puts out two different signals. It operates in the following workflow:

 

1. Relay 1 forces the projector to advance one frame. (or controls the motor to advance one frame, then stop with the shutter open. mechanical relay can lock the shutter open, if needed.)

 

2. Timer delays the rest of the system to wait for the mechanics to advance (with some extra time to allow for any mechanical varience. timer would be able to control motion if needed)

 

3. Relay 2 trips the digital camera to expose through the external shutter release, as any electrical connection between the two points in the jack will trigger a frame to be capture. (digital camera due to the better optics, resolution and picture quality availible, all at a cheaper price than a camcorder, esp a 24p camera with sync input)

 

4. Timer delays system for a little bit to let the camera expose. system resets to step 1.

 

5. optionally a binary counter can be incrimented with the first signal (one chip, simple design) and compared to a set number of frames (set by a binary encoder, a simple switch-like rotary knob that outputs binary. one, two or three switches, depending on the max number of frames you want it to take)

 

This design would be simple and require no programing skills (no microchip or computer involved), and building would be simple enough for a first timer to build. (schematics are very simple to understand, once a few basic concepts are outlined)

 

There are varibles in the design, depending on the projector in question. If it can't easily advance one frame on a relay closing, it could be designed several ways to control the motion of the projector, . It can even relay the light to come on, the action on the projector to pull down one frame, then when a switch is closed (indicating the shutter is open and frame is ready) the whole drive can be shut down by the timer, allowing frame control of the projector by the control box.

 

Then the device would trigger the D. camera to expose for as long as it takes. then repeat the cycle. It may take 1/2 second to a few seconds to take each frame (depending on the cameras speed) either way the timer can addapt. A slightly more complex system can by calibrated to stop after a preset number of exposures (allowing the operator to dump or switch the memory card if needed) and then resume at the press of a button. No frames would be dropped and the operator can dump those frames into premiere and render a file in any format they wish (including Hi-def if needed) frames would automatically be progressive with no pull down to make things easy. Also there would be no NTSC colorspace (quite the opposite, there would be full raw 10 bit pictures availible in any resolution the camera provides.) meaning higher quality color correction (premiere can add the NTSC color space and pulldown after color correction is done and the finnished cut is ready for export

 

I will post the design on monday or tuesday. I don't know if I will necissarily have all resistor and capacitor values listed, but I will try and post the formulas and some general discription of the design, and you can plug in the numbers based on how long it takes the projector to advance, the camera to snap etc.

 

Let me know matt what design requirements you need for your projector (if you can advance your projector with a relay closing or if you need a full control unit to turn the motor and to stop it when the shutter is open. let me know.)

 

It sounds like a complicated design, but trust me, any of the design required would take 1 chip, a few discrete parts (resistors, transistors and capacitors) and a couple relays and connecting jacks (like BNC connectors) all parts are redilly availible at any radio shack (though they over charge. There are specialty electronic stores that give better rates on parts and would be more helpful in store, since they are probably hobiest too. All these parts are very inexpensive) If you need the unit to pause after a certain number of frames, you may need 1 or 2 more chips to count and compare to a preset number, and switches that allow you to select numbers (1-16) and output in binary (also availible at radio shack.)

 

If anyone is interested I will even build it for them if they feel it is to difficult for them (I would charge a bit extra over cost of parts, since I don't even own a soldering iron anymore)

 

Also if you go to a specialty shop, be sure and talk to some people. There are some interesting people in those electronic shops. Some will straight up lie. I was trying to design a solid state audio player when I was very young (long before MP3s players or files and flash media made it easy) and was chatting with one guy who said he had 5 patents on an invention, and it was so delicate it had to be kept in a perfect vacuum. My urge to call BS on him was way too great, but i did resist.

 

Electronic Engeneering is not my mainstay, but for some reason after years of not designing anything, this has peeked my interest. I have been thinking of getting back into it, to design devices that could be useful on set for me(custom flicker boxes, Kino flow units, LED light pannels with the power of an 18K HMI etc)

 

 

 

an udpated thought when I read your macro suggestion, you can use a digital camea with the lens off and focus the light directly on the CCD and reduce the amount of light needed to expose (since light isn't reflected off a screen or passed through a ground glass. It is simply focused directly on the CCD of the DSLR, as if that was the viewing screen. This means a lens that can project the S8 frame the exact size of the DSLR's chip. Also since you have reduced the amount of air and glass, sharpness (like you said) and contrast and size would be reduced. Critical focus might be difficult however during setup, though once its set, you can leave it. Filters can be used for rudimentary 'timing'

Edited September 16, 2006 by Michael Collier

[Matthew W. Phillips]

Ok, wait a sec...this is getting heavy and I wanted to point out that the least amount of technical skill necessary, the better. I dont want to try to make a RANK. Just trying to come up with a frame by frame projector capable of some decent scans. I am not an engineer and i dont wish to become one. Any electrical work would have to be cheap, easy to understand, and safe.

[Michael Collier]

If you want, I will build it for you. you would only have to set up the projector's motor to the timer by adding a simple switch to the motor shaft or the shutter (send me pics of the motor shaft and shutter so I can design the switch for you) Then focus the projector on a screen with the DSLR focused on the screen, or the DSLRs chip. Then connect the DSLRs shutter release jack to the timers BNC connector (a simple wire cable I can make for you). Then all you have to do is enter how many frames to let it run before cards need to be switched out through the switches on the front and hit go. Let it run, change cards and hit go again. the box would take care of the rest. i can build it in a cheap box to make it totally safe.

 

 

you can even send me the projector and a proccessed reel and I will build the whole thing for you and test it. You would get better than rank quality images out of it for less than the cost of telecining a few rolls of 8mm (and still be able to use the DSLR to take pictures, when not scanning)

 

If you want to build the circut I can guarantee its within your technical abilities (schematics are just pictures that show the parts and wires that connect what part to where.)

 

All I would charge you is parts, shipping and an extra 15 or 20 bucks for a soldering iron and solder. I would do the work for free (its been so long since I have built something, I think it would be fun) I bet it could be built for less than 50 bucks plus a wall wart (I can design it to use the charger for your cell phone, or a wall wart you have lying around). Then connect two wires and hit go.

 

it might have a potentiometer knob to set the delay to allow the DSLR time to save. Set it too high and the scan takes a bit longer, too low and you would loose frames, so you would set it higher and try again. This would help if you used different DSLRs or different settings on the camera, affecting save time.

 

It would look something like this when built. 3 wires connects it up, 1 or 2 pots set up the timing. 4 switches sets how many frames to run un-supervised. two buttons to start and stop the device

Edited September 16, 2006 by Michael Collier

[Matthew W. Phillips]

That sounds good but I discovered last night that I still have a problem with the lamp i used not being powerful enough so I have to modify. Also, I didnt point this out earlier, but I dont have an electronic device advancing the film, it is a crank system. Would this fit in or have to be changed?

 

Right now, the only electrical part is the lamp which is a basic flood light mounted on the inside (wanted to make replacing lamps cheap)

Edited September 16, 2006 by Matthew W. Phillips

[Matthew Buick]

Yes, you can. Jaakko Kurhi (www.jkcamera.com) makes an optical printer that can accommodate regular 8mm, Super 8, 16mm, Super 16, and 35mm. I once owned an earlier version of the printer now shown at his website and with it produced reductions from 16mm to Super 8, straight copies, and multiple frames within a frame on 16mm. Jaakko also does conversions of H8 RX and H16 RX cameras to DS8. Great stuff!

 

How much will one of these "dangerous" machines cost ?

 

Optical Printer = Bad Ouchie

[Gianni Raineri]

I've studied the moviestuff way with my DIY & Dumpster Diving & how cheap can I go... I' dislike programming, electronics, and engineering... but admire the industrial design of cine film and the Projectors and Cameras. For guidance, inspiration, and luck to get that alchemy and creative muse flowing, I just pray to the 8mm goddess. B)

 

I find the cheap working projectors at flea markets (o.k. ... boot sales..., I'm in Britain). The last three projectors cost me two for .50p each, and one for £2.00, and they are in great working condition. OK one I thought it was broke untill I added drops of oil in all the main gears and joints. I recently improvised temporary "film chain of sorts" into a virtual quarz-chrome-o-scope quick time movies converter. I've done it previously with other projectors any recording device that works, cell phones, broken digital cameras, etc...

 

My latest system used a Sony Digital 8's slow shutter mode with negative art effect for Quartzchrome reversal film processed as negative. I removed the 50w bulb and replaced it with an modified Ikea 10w Quartz "NON" lamp (£1.00 bootsale find) and wow it works! Just this morning I bought an ancient Kodak glass color correction filter kit (0.50p) for testing transferring color neg film. I hear that orange mask needs a blue filter..... we'll see...

 

Sorry it's real time capture...it's not frame by frame which for me involves hacking a USB mouse (£2.00) and microswitch (£.50p) into the projector movement....to talk to the computer time lapse software. Even my experiments are only temporary and for just one movie, it's great for uploading personal stuff to my google video and youtube....my plans are to improve and get better as time goes by...

 

Here's the result, one and a half minutes as shot by my six year old boy, on a £2.00 super 8 camera, with outdated Quartzchrome reversal film, home processed in Ilford's D76 clone soup and telecined in the above system...

 

 

Why don't you try to adapt a transport from a motorised film editor, or a broken projector, or the guts of a melted cheese gear Kodak Instamatic, or other Super 8 camera to get the film transport working.

 

Sorry about any typo's in this message, I just cannot figure out how to edit these after posting it...

 

Gianni B)

[Gianni Raineri]

Here is the flicker pages with the NON-Quartz-scope quicktime converter, the other cheapo projectors to be used next, and my boy (the director of the above film) displaying the cheap plastic manual PLUS Super 8 camera he used on the shoot.

 

Goodimages on Flicker

 

 

Gianni in Londinium :lol:

[Matthew Buick]

Hey cool, I live in Newcastle, where the Millenium Bridge is, you'll see a lot of that in my films, I'm not passing up on a location like that. :D

Edited September 17, 2006 by Matthew Buick

[Michael Collier]

That sounds good but I discovered last night that I still have a problem with the lamp i used not being powerful enough so I have to modify. Also, I didnt point this out earlier, but I dont have an electronic device advancing the film, it is a crank system. Would this fit in or have to be changed?

 

Right now, the only electrical part is the lamp which is a basic flood light mounted on the inside (wanted to make replacing lamps cheap)

 

Mechanical advace would help, since a little automation would make the chore a bit easier (like ronco: Set it, and forget it!) That said, mechanically it would not have to run fast, so a high RPM, low tourque model could be geared down to advance the film. A machined roller can pull the film down, with a nylon roller or steel roller with sprokets above to keep the film aligned and tensioned. Then the switch can be rigged the roller to trigger the camera to snap a picture. since the camera takes a while to save the picture, the rollers can be slow. Keep the exposure on the camera set to a high shutter rate and you can ensure motion blur from the projectors motion would be mitigated.

 

this would simplify design, since the rollers speed would effectivley set the timing. a machinist can make a roller with the exact circumfrence of a few frames of film, with sprockets. (this roller can replace one in your current projector.) The rollers exacting dimmensions will make sure the films registration is perfect on scan. on one side of the roller is a plate with multiple blades to trip the switch (one blade for every frame the roller pulls down in one turn) The blade can be spun on the roller to adjust for any delays in the system, so the frame will be perfectly centered on the camera. Everytime 1 frame is pulled down, the camera is tripped.

 

If the camera has reliable delay between when the shutter release is tripped and actual exposure, then electronic design would be nill, since the switch can be rigged directly to the shutter release via cable. If the camera doesnt reset before the next frame, simply slow down the motor by changing gears or reducing power to the motor. Most DSLRs have a very constant time between tripping the shutter and actual exposure

 

That would be the simplest setup. Below i will outline a few more options, but keep in mind the would be more design intesnsive (I am still willing to build it) but could end up saving money. Machinist are expensive, though I think rollers with sprokets would be very easy to machine, taking 1-2 hours. Maybe a roller with perfect dimensions already exist. in S8 1 sprocket = 1 frame right? If thats true then any sprocketed roller ever made by nature would have to have the right dimmensions.

 

If machining a roller is too costly, then the only alternative is to put a switch on the motor. Then calculate how many times the motor has to turn to pull down exactly one frame. Take into account the circumfrence of the roller and the gearing (preferably a worm gear on the motor, and a large gear on the roller. The higher the gear ratio, the better the 'registration'), to determine the exact count. From there I can build a circut that can be used to count the motor rotation, until 1 frame has advanced, and trigger the camera release. A calibration knob control would have to be built in, to control the frames alignment (sort of like V-sync on a TV.

 

 

The motor can speed up and slow down without any problems in registration. if the camera in question is does not accuratley snap a picture based on the shutter release (High end models will have a short, constant delay. Cheaper ones may have more variations) the design can include a motor cut off to stop the frame for a moment while the camera takes the picture.

 

there are tons of simple ways to do it, even with your setup. I guess I just need more specific pictures (and higher resolution) to tell what method would be best for your design. I am thinking about building an S8 home HD telecine to sell, if there is enough interest in it. It seems like there are tons of people still shooting S8, but getting it telecined is getting more expensive.

 

 

Let me know though, if you have a roller already build in then all you need to do is add a switch connected to the shutter release, and a slow moving motor to turn the crank for you. It could be as cheap as 15-20 bucks to get this all set up for you, and easy enough for you to build, not being technical and all (if you can wire a lamp, you can rig this. shutter release for DSLR is really just a switch hooked up to a mini-jack)