Homemade Telecine

[Stan Jelavic]

Got the ML (MagicLantern) firmware running again in my Mark2. It is slow but produces huge images
in RAW format. Obviously they can be resized to whatever size you want.

Reused Wolverine reel mounts for now.
https://photos.app.goo.gl/eJ13ALUTUyu6g7e2A

Checked the MSP430 pinouts and looks like we have enough to accommodate the new functionality:
MSP430 pinpouts
================
P1-0 Step
P1-1 Takeup out
P1-2 Test up
P1-3 EN
P1-4 Test down
P1-5 Align switch
P1-6 Rew input/16mm in run mode
P1-7 Run input
P2-6 S8/R8 switch
P2-7 Trigger out

Added the test mode. The Test up and Test down buttons can be used to change the roller cadence to match the frame length in test mode. The test mode is entered when the Test switch is activated.
When the test switch is deactivated the new cadence value is stored in the flash.
Added the S8/R8 switch. Note that the test mode will have 2 cadence values stored depending on the position of the R8/S8 switch.
Possibly three values if 16mm is used.
The 16mm mode is controlled by the Rew switch.
Actually it may be possible to have the 35mm mode instead of 16mm.
35mm could be useful to scan the 35mm negatives.
Obviously all of this has to be properly tested. Quite a bit of work indeed.

[Stan Jelavic]

Finally managed to write a few bytes into the flash memory of the MSP. With that working I have to come up with the tuning code. Here is a rough outline what I have in my mind:
Turn run switch on
Turn test switch to on. This still runs the main stepper code but also enables the test code that is run after the main code.
The test code polls for up and down buttons
If up button is pressed the stepper delay gets decremented by 1. If the down button is pressed after that (while holding the up button down) then the fast decrement is activated and the stepper delay gets decremented at the rate of 10 per second)
The decrementing stops after either button is released.
Similarly, if the down button is pressed the delay increment is started.
Turning the test switch off writes the delay into the flash.
It is to be noted that the flash gets read upon every power up.

[David Brown]

Wow I love to see the last picture that includes the supply and takeup reels. I most relate to you last words, "Quite a bit of work indeed."

I am getting stronger every day, but I still relate to things that take a lot of work!

I have an idea to build rewinds that can adapt easily from 8/S8/16mm. 16 has a square hub that's close to 8mm. If the pinch roller were wider it could cover all 3 formats leaving 3 interchangeable gates. I'm feeling tired already!

35mm. Well, that would just need a gate. I have a film scanner, software and film holders. I have been considering using a DSLR instead. I bought the scanner 10 years ago and have not fully begun to scan negatives. My scanner is still on the market and I doubt they will see any resolution increase.

I'm rambling and you could be joking about 35mm!

[Stan Jelavic]

I guess get the R8/S8 working first. I mentioned the 35mm because I had lots of 35mm negatives and it took me ages to scan them in. Not sure if it would be any faster with the homemade scanner because you would still have to align the film.

[David Brown]

I think the camera would be faster. The scanner is always in focus, but then there is prescan to crop then the real scan. Mine is very slow! That said there is software to scan a whole tray and save separate images. Why not have the option here? You might need a larger led lamp?

I'm printing a 16mm takeup reel right now. I have a metal reel of 16mm but the film on it might all be black? Won't know until I can wind it on another reel.So a 16mm reel is the same as an 8mm reel, even the key. It is thicker than 8 so it won't fit on the wolverine without some way to lengthen the spindle.

I also have a reel of 35mm, it is microfilm and the reel uses the same 16mm hub, but that's of no use as 35mm negs will be short even if uncut.

I'm thinking of a single gate that has inserts for S8, R8, & 16. When empty It could hold 35mm film or even a slide?! The pinch roller with two of the rubber rollers fits the 16mm width. I'll go there later. I only have some early prints of the gate parts. I was waiting until you had enough time to make changes based on usage.

I can work on the alternate gate now. Hey, how about a gear to advance the 35? No sprocket damage.

[Stan Jelavic]

One issue with the large gate is the light distribution. It has to be even across the gate. I am trying to get the new MSP firmware working with tuning included. Got the code done but have not tested it yet. Once that is done I will do a test clip and make sure that the film shake is still within bounds and then complete the new board. The new board will have the tuning mode. After that will look into 16mm and possibly 35mm.

[David Brown]

Yes, I don't get much done in a day!

I did print the 16mm takeup reel. Surprised to find half of the film is kodachrome and the rest is B&W. It's a 7" reel, 400', so there is probably 200' on the reel.

[Stan Jelavic]

Use Launchpad to debug the tuning firmware.
https://photos.app.goo.gl/UqUHzfSsDEESTXVZ7
Due to the complexity of the code it is necessary to run the Code Composer debugger. The three resistors are the pullups on:
- tune switch
- tune up
- tune down
It all works now. Here are the tests:
- tune on
- toggle up switch several times and the count goes up
- tune up switch on first then down switch. after a minute the count went up to 1000.
- repeated for down mode
- tune switch off, the count gets written into the flash
- power the board down and up, the count gets read back from the flash correctly

The next steps:
try the new code in the target
add r8/s8,16mm tuning

[David Brown]

That's a lot of steps. I'll guess more steps equals more speed. I thought I'd send you a length of 16mm. That won't help until there is a gate, not to mention the roller. I have modeled the 3 films and began with 35mm thinking I'd work my way down. I thought that 8 & S8 could use the same gate if there were two apertures in the same gate. Going slow, every day I feel a little better!

[Stan Jelavic]

Actually it is not that bad. Check if sync is too short or too long. If it is too short:
- turn tune switch on
- turn up switch on
- turn down switch on and observe the sync getting longer and longer (longer stepper runs)
- Once close turn up and down switches off
- Now using the up and down buttons (yes there are up and down switches and buttons) get the stepper to sync up
- turn tune switch off. This stores the new value in flash.
- recycle power

If the sync is too long then the down switch on first followed by the up switch. The rest of the procedure is the same.

The next time you use the scanner you do not have to go through the above procedure. But if you notice after a long run that the stepper is getting out of sync, then turn the tune switch on and then adjust the stepper by using the up and down buttons on the board to tweak the sync.

Tried this on with the scanner and it does work. My problem now is with the Magic Lantern. It times out on me. I get a few clips in and then camera clicks out of the Live Preview. Will have to study the forums.

[David Brown]

That sounds great. I get it now!

All cameras time out on live view, I think it has to do with the battery and a preset sleep timer. Good luck.

I have an awful time sketching with a pencil, so I have been sketching in 3D. These pictures were built by eye and not precisely measured (except the films). It has been a little slow just making cubes, stretching, and cutting holes. Better for me when visualising everything. I could print this, but it will be sloppy and there is more to do.

I don't know how to plan around a light source yet. The box could be smaller. The goal was to keep each film centered so minimise camera movement between formats.

Let me know if anything grabs your attention.

[Stan Jelavic]

The issue is not with the Liveview timeout. It is disabled. I believe it is a glitch on the trigger pulse that may be causing the issue periodically. Will have to hook up a scope to it.
The transport works pretty good.
https://photos.app.goo.gl/PC2HaA23sbnENSbb7
There are some issues that I noticed. The gate has some fluff or something in the right hand corner. Will have another look. It creates like a cloud shape in the image.

The scan process is very slow due to MagicLantern raw image processing.
The light source is not bright enough causing long exposures.
The gate pressure has to be set right in order to reduce the amount of film shake. Too much of film shake results in a
large amount of cropping.
Here is the final cut.
https://drive.google.com/file/d/1CcX...ew?usp=sharing

Have not spent much time on your sketches. Looks like you have some very good ideas there. Will have a closer look tomorrow.

[David Brown]

Stan, do you have the a link to the white rubber roller? There is only one in the pictures below.

I think that just the same width rollers can move 16mm film. I tried to test that and remembered that the wolverine board is now defective.

Two rollers could be used side by side. The Rocker would need to be widened. I had to remove the circular ends to clear the wider 16mm film. I'm also using some dual 8/16 rollers I found on thingyverse. They are too big, will need to make my own.

I also found a 35mm roller (with teeth). It can also be made smaller, but if it is to be motorised it will need some more thought.

[Stan Jelavic]

Here is the link David:
https://www.fruugo.us/recorder-parts...2325-124662196
I still have some left and can send you a few.
It should work with the 16mm film if two are mounted side by side and the rocker redesigned to accommodate two of them.
The toothed roller should also work with this controller.
Had an idea how to mitigate the film shake without cropping the border. Just open up the gate to get a larger image. With that the shake will be outside the image area. Pretty simple. This makes the design simpler i.e. allows for more shake that gets fixed in the post.

[Stan Jelavic]

Checked the gate under the microscope to see what the cause is of smudging around the edges.




As the images show, the cause is a few burrs and rough edge from the printing process. Not a bit issue. Can be easily
fixed.
Actually I think the gate opening should be wider to allow for larger overscan. That way if the film shake is excessive the edges can be trimmed
without cutting into the image area.
Now to the next issue.