DIY 16mm Film to 4K Digital Transfer System | Part 2

Here's Part 2 of my Do-It-Yourself 16mm film transfer system. I've modified an Arri 16S camera movement for pin-registered, high quality film transport. In this video, we'll lay out the components, attach the film mechanism and start the basic build. The goal is to produce a system capable of frame-by-frame 4K transfer of negative and positive Super 16 and Regular 16 film.

[@JacobCarlson]

Excited to see the stinkbug cameo

[@flyingo]

Fascinating and inspiring. As a former television station chief engineer, I’m constructing gadgets and mounts for my odd ideas quite often. Now retired, I’m back to the inner filmmaker of my youth, so I’m eager to see how your project comes together. Thanks!

[@brianmuhlingBUM]

This stuff gets better and better every episode. I cried to see you dismantling the Arri, but I guess it is for a good cause.

[@FreshGroundPictures]

Please don't shed any tears - I bought the Arri parts as they were, salvaged from cameras that had already been trashed from a college film department. I hope you get to watch part 3 of the series, as the final result turned out even better than I expected!

[@SHDEdits]

Awesome build. If you replace out the white LED you're using with an RGB LED array, you'll be able to improve the signal to noise ratio of the capture by neutralising the orange mask with a bluish output from the array instead of adjusting the analog gain (white balance) of your camera. That will maximise code value resolution for higher densities. For densitometric calibration (assuming 16-bit radiometrically linear raw output is available) adjust the output per RGB LED and exposure time such that d-min sits at an average code value of approximately 42000 per channel, this equates to a transmittance of 0.640 or density of 0.193 which will offer plenty of bottom room for D-min granularity. The older stock you're scanning has an appropriate density range for a traditional cineon log encoding. Assuming you've met the calibration conditions stated above for D-min, whatever the 16-bit linear code values are for each pixel going forward, divide it by 65535, this will solve for the transmittance of that pixel, convert that pixel transmittance to density via -log10(transmittance), then convert to 10-bit code values via 95 + 500 * density. This will produce cineon log code values in scanner density, printing density will require further calibration.

[@curiosity2314]

So far it looks very good!

[@RMphy89]

Great content! I am in the process of restoring a Victor Ultra Cine 16mm camera from 1924, so seeing this project and its reuse of camera components is fascinating.

[@FreshGroundPictures]

Thanks - good luck and have fun!

[@Snusmumrikus]

Nice, been looking forward to seeing more of this project!

[@FreshGroundPictures]

Awesome, thank you!

[@TheCinematographyLab]

Awesome work!

[@FreshGroundPictures]

Thanks a lot!

[@sparky60ful]

Make your own flatbad! Great video again! Arri has used M2.3 and M2.6 screws in the past.These where sizes that could be bought in a good hardware store about 30 years ago for electric appliances. But non anymore, sadly. Now we only have M2.5 in between M2 and M3 left over. I still own the taps and dies for these rare screw and nut sizes.

[@FreshGroundPictures]

Thank you for the information- very good to know!

[@johnturk8622]

Excellent video! I am in the process of building an Optical Printer with a Bolex REX 4 as the camera and something very similar to your machine as the “Projector”. The clarity of your presentation is a real inspiration. Thanks.

[@FreshGroundPictures]

You might already be aware of this, but years ago there was a device made by J-K Camera Engineering that was an optical printer for 16mm - it had a 16mm "projector" with a registration pin, and you could attach your own camera. It was set up on something like a lathe base for stability and adjustment. I would have loved to find one of them! Thanks - good luck.

[@johnturk8622]

Yes indeed, the JK optical printer is a device that I am very familiar with having spent countless hours using one in the 90’s. They are all but extinct and SUPER rare these days. With your acumen for engineering and passion for all things film, you could probably easily build one yourself. They were amazing machines that were capable of very creative work. Thanks again for your thoughtful presentation in your work. As all the others who follow you, I look forward to the next installment. Cheers.

[@erchata]

me ha parecido muy interesante ( me he suscrito) un saludo desde BARCELONA ESPAÑA.

[@STARanoff]

Amazing!

[@FreshGroundPictures]

Thank you! Cheers!

[@tbip2001]

Brilliant . I wodner if all these components could be drawn and 3d printed to make a ready msde kit? Do you think there’s any way to bypass the whole claw/movement, and just jave a the film moved the exact amount by pulling it through with a precosion stepper motor?

[@FreshGroundPictures]

It's probably possible, but without registering the film in the gate, I think you'd have to use some type of software to align the sprocket holes digitally. I think Moviestuff machines do that. I'm more mechanically inclined (not an expert though) so I felt that a hard gate and registration pin would provide stable, sharp and flat results from the start.

[@maxdoes_]

The build and video are both incredible! Can you share some more info about the adjustable camera bracket? I'm looking for one just like it for a project but can't seem to figure out the right keywords. Thank you!

[@FreshGroundPictures]

Bought it on Amazon - Koolehaoda 240mm Professional Rail Nodal Slide Metal Quick Release Clamp - $34

[@maxdoes_]

@@FreshGroundPictures Thank you!

[@heatonize]

8:16 i got so scared when the bug came on screen

[@FreshGroundPictures]

Literally had a bug in the system!

[@MaxKissler]

I'm curious to see where this is going. Does it work without any sprocket wheels and loops next to the gate? Doesn't the alignment change due to the fact that the takeup spool gets bigger and therefore a constant motion by the motor misaligns or stresses the film? I'm about to built my own 16mm Scanner out of an old projector using an LED and a stepper motor. At least it should have all the necessary mechanical components already in place for an easy starting point - two sprocket wheels in addition to the film registration mechanism in the gate and the spools all connected by one gear system. Ideally I'll be able to wetgate scan by adding a water reservoir and a pump and some blower fans to it.

[@FreshGroundPictures]

Yes, excellent observation! My original idea was to just use the take-up and supply spindles as I've shown - they have slip clutches built in so that if the tension reaches a certain point the clutch slips and stops pulling momentarily. In theory this could work for a whole roll of film, however I now feel that the tension gets a bit too high on take up. As a result, (and I will show this in a video) I'm putting in drive sprockets from the magazine that will allow a loop to be formed so that there's always even tension on film entering and leaving the gate - as I said, it's getting a little more complicated. The wet gate is a great idea, but I thought that a particular oil is used that matches the refraction index of the film base, and dries quickly? Thanks for watching!

[@MaxKissler]

​@@FreshGroundPictures Nice. Can't wait to see the results! I guess I should be looking into the whole wetgate scanning process more thoroughly. I thought of just using distilled water with a wetting agent like the ones used in the final stage of black and white processing. I fear you are right though, it's probably some oil. At least I remember a friend who runs a drum scanning service mentioning oil and what an annoying process it to clean the negs afterwards.

[@AfricanTvFilmaker]

I hope you finish the prototype and sell it to a manufacturer, I would buy it, especially if it had an Ultra 16 gate.

[@roybrown503]

Was there a reason why you didn't adapt a 16mm projector as you did with the 8mm project?

[@FreshGroundPictures]

Because 16mm is mostly shot professionally with a tripod, I felt that image stability right into the camera was very important, hence the pin-registered film gate. There are a lot of projectors out there, and some might offer similar stability if not worn too badly. Rather than testing a bunch of projectors, I had the film camera parts, so I thought it would be fun to see if I could make them work, in essence "reverse" the process and still have the stability and film handling safety of a camera.

[@darklord_ice]

could you please a buying source / the exact designation for the LED? thanks

[@FreshGroundPictures]

I'll get back to you - the LED unit I used is no longer made, and I have to check if there's a replacement. I think it was 95 CRI - the higher the color reproduction the better.