Hacking the Kodak Reels 8mm Film Digitizer (New Thread)

larryc39

New Tinkerer
Jan 2, 2025
32
24
8
Do NOT try to eliminate the possibility that rollers give scratchmarks on the filmbase USING heat-shrinking tube. I did a test, se pictures, with transparent tube. The surface is unfortunately to soft and will give friction when the film is somewhat stretched. Maybe if one uses another heat-shrinking tube, like black.

I would not try and use heat shrink tubing, it's pretty soft and will add friction unless you can find some of the glossy stuff. Also you'll ideally want something that is loose and rolls on the posts like a pulley, not staying stiff and immobile like the current posts. Someone has modded theirs to include nylon coverings on the posts that rotate and someone else has use PEC tissue or Teflon tape on the existing fixed posts to help reduce the friction.
 
  • Like
Reactions: Claus

videodoctor

Tinkerer
Jan 8, 2026
78
30
18
We now have to get the bit-rate up on the Type D, as you have achieved more resolution, but too much is lost to compression.
I'd like to be working on this effort---I brought up before, but it sounds like it's not a good idea to have a temporary "fix" that just hard codes the Qp value as it will crash the encoder according to @0dan0. In his notes that he posted a while back (a TXT file with some of his changes documented):

In Memory (I've found)
0xa56f1f60 - Current Qp level for the H264 encoder

In Memory (I create)
0x85bf001C - Current Qp

0x23fe00 - used to be a call to 0x241be4, now it is a hooked call to 0x1fc280 (sets and prints to initial Qp value), which still calls 0x241be4.

@0dan0, would you be able to outline what general steps I could take with Ghidra analysis on Stock D to get a short term bitrate fix implemented?
 

0dan0

Active Tinkerer
Jan 13, 2025
424
567
93
I'd like to be working on this effort---I brought up before, but it sounds like it's not a good idea to have a temporary "fix" that just hard codes the Qp value as it will crash the encoder according to @0dan0. In his notes that he posted a while back (a TXT file with some of his changes documented):

In Memory (I've found)
0xa56f1f60 - Current Qp level for the H264 encoder

In Memory (I create)
0x85bf001C - Current Qp

0x23fe00 - used to be a call to 0x241be4, now it is a hooked call to 0x1fc280 (sets and prints to initial Qp value), which still calls 0x241be4.

@0dan0, would you be able to outline what general steps I could take with Ghidra analysis on Stock D to get a short term bitrate fix implemented?
I did not use Ghidra all that much. From my memory there is an error message for when encoding is struggling with a high bit-rate, I use this to control the Qp. When I saw the warning, I increase the Qp 3 or 6, depending on the warning level. Set the Qp to 10, then using grainy film, that watch of the err message for the encoder buffer overflowing (they aren't that literal.) This look for where the message is printed out, and stub that code, to reduce the bitrate (increase Qp).
 

ThePhage

Tinkerer
Oct 30, 2024
62
59
18
Hi @ThePhage , I am zoomed all the way out. I have the sensor mounted 7 mm higher than original. If I move the frame some clicks(2-3)) up or down, the pictures edge will be visible.
I have not tried to scan with the lens mounted in other distances due to what was written in the Great Manual for modding the scanner. So I thing is is fairly optimized, don’t You.
One of the features of 0dan0's modified firmware (which will hopefully be forked to Model D scanners) is the ability to digitally Zoom out wider than the stock firmware supports. This way, if you desire, you can scan the entirety of the film frame and then crop it precisely in post/editing. I do this since my old footage often shifts around for various reasons. I scan wide-ish, and then re-frame each clip in Davinci Resolve.

Additionally, as 0dan0 just mentioned above, the custom firmware vastly improves the encoding with a much higher and controllable bitrate, and many other tweaks to improve color handling, exposure, frame rate, etc. For more, see the link to the user guide in my signature.
 
  • Like
Reactions: videodoctor

ThePhage

Tinkerer
Oct 30, 2024
62
59
18
I would not try and use heat shrink tubing, it's pretty soft and will add friction unless you can find some of the glossy stuff. Also you'll ideally want something that is loose and rolls on the posts like a pulley, not staying stiff and immobile like the current posts. Someone has modded theirs to include nylon coverings on the posts that rotate and someone else has use PEC tissue or Teflon tape on the existing fixed posts to help reduce the friction.
After some use of wrapping PEC tissue around the existing posts, I've decided against it because after some time, individual fibers from the tissue shed off and then ride on the film and get stuck somewhere in the gate, ruining a bunch of scans. I've moved to a similar wrapping of the posts with pieces of fine cotton. Not the smoothest, and probably adds some amount of friction, but I'm not confident in switching to rollers.
 

lrdgreen89

New Tinkerer
Jan 12, 2026
3
0
1
After some use of wrapping PEC tissue around the existing posts, I've decided against it because after some time, individual fibers from the tissue shed off and then ride on the film and get stuck somewhere in the gate, ruining a bunch of scans. I've moved to a similar wrapping of the posts with pieces of fine cotton. Not the smoothest, and probably adds some amount of friction, but I'm not confident in switching to rollers.
What about simply wrapping scotch tape around the posts? (provided the edge of the tape isn't where the film touches)
Also, is the PEC pad still recommended for use in cleaning, or is there a risk of depositing tissue?
 
Last edited: