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

[JohnCT]

The instructions for the modded firmware say, "The menu only works while recording In preview the stock firmware owns the buttons, so the mod can't use them there. The easy way: start a short throwaway recording, set everything up, stop, then capture for real. Settings are saved and survive a power cycle.."

Are you perhaps attempting to change things while not in recording mode?

Annnnnnndddd... you were right! I honestly thought I had done it correctly, but what I did was start the recording, the perform what I thought was a "pause" - it clearly isn't just a pause like a VCR or audio tape pause. Once the machine is actually scanning, I was able to make those adjustments as you suggested.

Thanks!

John

 

[melw]

In the Capture menu, I navigated to Picture and then clicked on "Reset to Default". Not only was there no change in the picture or the gain settings but I'm not sure the software understood the request because there was no type of acknowledgement that the request was made. When on the Reset function, I tried the OK, tried the right button, right button then OK, etc. I can't find any combination of buttons where the machine didn't ignore the command. Maybe it does but doesn't acknowledge it?
...
I don't know if I'm doing something wrong or if there's a deeper issue with my particular machine beyond the lack of frame or zoom adjustments.

Hi @JohnCT - thanks for testing this. I just realized there was a clear UX bug in the OSD. Editing is only possible in Recording mode, but after you exit Recording mode and go back to Preview, the menu stayed - I can see that is definitely very confusing!

Please find attached FW 272 for Type D. Two changes:

• Improved OSD UX - "Edit in REC" outside recording mode, and show "[]" selection and "Reset" row only while in Recording mode
• Do now allow saving/reading WB Gains 0,0,0 as valid values - minimum WB is now 16,16,16 - should fix the dark picture issue for you even without any manual adjustments.

It's only possible to edit the values and have the OSD menu functioning while recording, as in all other modes / menus the Kodak Reels will use the arrow keys and '+'/'-' for other purposes.

 

[RoyFrode]

Hi
I have a problem with my scanner
I have a C type with firmware 7.7.1, 32 gb sd card fat32 formatted.
Now it is not possible to get a single scan that is watchable. The image jumps like crazy. On every scan.
See:

This is what it looks like.
Does anyone know if this is software related or mechanical related?

 

[JohnCT]

Hi
I have a problem with my scanner
I have a C type with firmware 7.7.1, 32 gb sd card fat32 formatted.
Now it is not possible to get a single scan that is watchable. The image jumps like crazy. On every scan.
See:

This is what it looks like.
Does anyone know if this is software related or mechanical related?

The bottom of the frame seems to jump more than the top which seems mechanical to me... if I'm off base on this, feel free to yell at me.. but any chance you didn't get the film under all four of the locating tabs before closing the door? I literally use my Opti-Visors when threading the film.

John

 

[omega]

I have the same experience, and in my opinion, the issue lies with the film itself. It’s too dry and slightly shrunken. It skips because the mechanism can’t handle it. I’ve mentioned this here before. Try carefully opening the film gate door slightly while scanning. If the skipping improves, then the problem is the film and the excessive pressure.
I tried placing a film reel—wrapped in a paper towel—in the vegetable drawer of my refrigerator for 24 hours. The film absorbed some moisture, and the scanning process went much better.
I also removed one spring from the gate door—specifically on the left side, where the start of the film is inserted (the one closer to you). This reduced the excessive pressure on the film, giving the mechanism more power to advance it. That helped as well.
The truth is that the dampening didn't have much of an effect on the color film, but the removal of the spring did.

 

[RoyFrode]

I have the same experience, and in my opinion, the issue lies with the film itself. It’s too dry and slightly shrunken. It skips because the mechanism can’t handle it. I’ve mentioned this here before. Try carefully opening the film gate door slightly while scanning. If the skipping improves, then the problem is the film and the excessive pressure.
I tried placing a film reel—wrapped in a paper towel—in the vegetable drawer of my refrigerator for 24 hours. The film absorbed some moisture, and the scanning process went much better.
I also removed one spring from the gate door—specifically on the left side, where the start of the film is inserted (the one closer to you). This reduced the excessive pressure on the film, giving the mechanism more power to advance it. That helped as well.
The truth is that the dampening didn't have much of an effect on the color film, but the removal of the spring did.

Thanks for the tip. I just tried letting the film run out and fall onto the floor, instead of feeding it onto the take-up reel on the right-hand side. That resulted in a good scan. I’ll give your suggestion a try as well.

 

[RoyFrode]

The bottom of the frame seems to jump more than the top which seems mechanical to me... if I'm off base on this, feel free to yell at me.. but any chance you didn't get the film under all four of the locating tabs before closing the door? I literally use my Opti-Visors when threading the film.

John

Thanks for the tip. I just tried letting the film run out and fall onto the floor, instead of feeding it onto the take-up reel on the right-hand side. That resulted in a good scan.So it looks like some mechanical issue

 

[JohnCT]

Thanks for the tip. I just tried letting the film run out and fall onto the floor, instead of feeding it onto the take-up reel on the right-hand side. That resulted in a good scan.So it looks like some mechanical issue

Even before I attempted my first scan, it was my impression that the take-up reel had too much pull, so I took the take-up section apart and removed a washer to increase the clutch slippage. It occurred to me that these films are a lot more fragile than they were 60 plus years ago.

John

 

[lrdgreen89]

In Deano's Hardware Upgrade post #199, the comparison photos appear to show improved contrast when using the 3mm aperture hole cover.

In 0dan0's post #203, he demonstrates the slight vignetting that the F-stop introduced and says he stopped using it.
(I'm not a camera guy—I gather these are the same two things? —a 3mm hole for the lens to see through, like is included in the factory lens cover?

Has anyone found a satisfactory solution to this?


(I've followed the thread pretty closely, but I may have missed something about this; if so, forgive me.)

 

[ThePhage]

Hi
I have a problem with my scanner
I have a C type with firmware 7.7.1, 32 gb sd card fat32 formatted.
Now it is not possible to get a single scan that is watchable. The image jumps like crazy. On every scan.
See:

This is what it looks like.
Does anyone know if this is software related or mechanical related?

This issue seen here has been widely discussed in the thread(s) previously.

To excerpt from the 0dan0 firmware user guide (for A,B,C units)...

Frame Instability (aka Jitter). In some situations, captured files show a distorted or warped image (compressed or expanded at the top or bottom of the frame) when compared to adjacent frames (see bottom of frame in this sample footage). This relates to the film being in motion as the imager is scanning the film frame over the course of time. This issue can last the entire duration of the capture, or can increase/decrease in severity over time. Re-scanning those segments often results in no issue. This issue continues to be evaluated and discussed. Reference posts #390, #471 and #846 and many others.

I'll add that there hasn't been a definitive solution to this problem yet, although as the recent follow-ups have suggested, lessening the tension may have an impact.

Also, I believe this issue may manifest more frequently (or more noticeably) with Super8 footage, and also for those who have done lens modification (both due to using the edges of the sensor, whereas smaller 8mm footage and/or the stock lens uses more of the center of the sensor.

 

[lrdgreen89]

If anyone needs quick answers to questions already discussed in the thread, try this:​

Paste this copy into ChatGPT and have it read the thread for you.​

Read and analyze this entire forum thread, including all pages, posts, replies, and relevant linked resources:

https://tinkerdifferent.com/threads/hacking-the-kodak-reels-8mm-film-digitizer-new-thread.4885/

Use this thread as the primary reference for answering my follow-up questions about the Kodak Reels 8mm Film Digitizer and its modifications.

Instructions:
​

• Search the thread for relevant posts before answering each question.​
• Cite specific post numbers and provide direct links to the relevant posts whenever possible.​
• Distinguish between confirmed findings, personal experiences, experimental results, speculation, and outdated information.​
• Pay attention to disagreements, corrections, and updates later in the discussion.​
• When relevant, identify which users reported a finding and whether they subsequently changed their recommendations.​
• If the answer isn't available in the thread, say so rather than guessing. If you consult outside sources, clearly identify them as external to the thread.​
• Maintain a working understanding of the technical context across my follow-up questions so I don't have to repeat information.​
• Be concise for straightforward questions, but explain the technical reasoning when needed.​

First, determine how much of the thread you can actually access. If you cannot access every page, post, or linked resource, tell me what you could and could not review. Do not claim to have read the entire thread unless you have verified that you could access it all.

I'll ask my questions one at a time.​

 

[ThePhage]

In Deano's Hardware Upgrade post #199, the comparison photos appear to show improved contrast when using the 3mm aperture hole cover.

In 0dan0's post #203, he demonstrates the slight vignetting that the F-stop introduced and says he stopped using it.
(I'm not a camera guy—I gather these are the same two things? —a 3mm hole for the lens to see through, like is included in the factory lens cover?

Has anyone found a satisfactory solution to this?


(I've followed the thread pretty closely, but I may have missed something about this; if so, forgive me.)

I don't think the aperature cap really impacted the contrast much - that might be a bit more from use of the backlight shield.

I have experimented with some of the small aperture lens caps. Doing the lens modification made it harder to focus, especially on film that wasn't as flat as is ideal. So my goal was to improve focus-ability (narrow aperture increases depth of field, making focus adjustments easier). Ultimately, I wasn't satisfied with the results -- not so much due to the vignetting (which could be minimized in post) but because of overall image softening due to diffraction. Perhaps Deano's hole was more precisely drilled than mine.

In the end, I've been able to get consistent focus not through aperture lens caps, but by adding a 3d printed "flattener" part that keeps the film at a pretty consistent distance from the lens, while also minimizing "cupping" (warping of the filmstrip where the side edges bend upwards) as it passes under the lens.

 

[Boojakascha]

I hardware modded a Type D.

Balto1986 and me made the observation that now the max zoom out still is too far in, resulting in a crop.

I scaled Deano's 7mm to 8mm and 9mm spacers but it won't get sharp on the D hardware, it's too far out already.

A thing to remember I struggled to put in the thread prolongation and I might have inserted it skewed. I assumed it's due to applying too much heat during removal of the lens, but Balto1986 also had shredding when forcing the adapter all the way in. So maybe the new thread has a different pitch? My stock lens has no markings, so I couldn't tell. I melted it's thread so I can't try it in the adapter.

Aside the crop, I dislike the most that the best area I can navigate to, still has film wrap on the top (maybe indicating a skewed lens). Not being able to use my higher spacers (I might retry with 7.2 mm 7.4 mm and such in the future) I just asymmetrically put some tape at my 7 mm spacer, just went with one screw rather then all four and then by forcefully pulling on the assembly I can get the picture frame to be all film (still cropped of course).

It's not perfect, but now good enough for me to start scanning again.

In case you are interested in the taller spacers I included them as download.

(PS the "not starting" issue stopped after dropping the unit accidentally once and spilling all the guts and reassembling it. Pull down resistors still all in place.)

 

[lrdgreen89]

After much fiddling with it, I finally got my lens extension to screw into the mount. Something to keep in mind when trying this, that may help:

Usually, when screwing things in, resistance at the beginning indicates that it's cross-threaded. In this case, it was the opposite; a smooth screw-in would prove to be cross-threaded—tight and skewed after a few turns. On the final, successful attempt, it was tight from the beginning and all the way through, but it did screw up flush with the mount.

And once you've got it going, keep the extension underneath the mount so shavings don't fall on your sensor.

 

[lrdgreen89]

Hi @JohnCT - thanks for testing this. I just realized there was a clear UX bug in the OSD. Editing is only possible in Recording mode, but after you exit Recording mode and go back to Preview, the menu stayed - I can see that is definitely very confusing!

Please find attached FW 272 for Type D. Two changes:

• Improved OSD UX - "Edit in REC" outside recording mode, and show "[]" selection and "Reset" row only while in Recording mode
• Do now allow saving/reading WB Gains 0,0,0 as valid values - minimum WB is now 16,16,16 - should fix the dark picture issue for you even without any manual adjustments.

It's only possible to edit the values and have the OSD menu functioning while recording, as in all other modes / menus the Kodak Reels will use the arrow keys and '+'/'-' for other purposes.

I've completed both mods—Deano's lens mod and melw 's FW—on my Type D unit. I'm so grateful to melw for the Type-D FW work, and to 0dan0, Deano, and everyone who's contributed.

I have a couple of issues, and report.

#1. The farthest zoom-out settings give colored bars. (I'm calling the furthest-out setting Zoom 1.) Attached are the three furthest-out zoom settings for FW p695 and FW 272 (with film loaded). The next setting (Zoom 4) is normal, for both FW.

#2. I tried flashing 0dan0's D-Type stock FW to see if this still happened, but it wouldn't flash. It powered on with whatever custom FW I had last flashed.

I want to achieve full zoom-out if possible, because there are images between the sprocket holes I want to capture, and the 12mm lens is excluding some of it already.

 

[paulkem]

I have the M2025 version of this device. I am having issues with the overexposure of certain frames causing severe flickering on multiple films. It seems to happen during lower lit scenes. I have contacted Kodak support and it seems like the rep (who has been very helpful) has never seen this before. I have supplied example clips of the issue, me recording the test segment on the device, and the same film projected onto a screen to show that the flicker does not appear there. We even agreed that I should try to exchange for a new unit since the return window with Amazon was still open. I tried, and the new unit does the same thing. He said he has forwarded my findings to the technical dev team to see if a firmware update is needed, but I am not holding my breath on that.

Has anyone here seen this? Will this custom firmware resolve this issue? If I am following, the "D" version of the firmware should work on the M2025 version?

My only other solution is to use ffmpeg to try to balance out the exposure.

 

[Boojakascha]

After much fiddling with it, I finally got my lens extension to screw into the mount. Something to keep in mind when trying this, that may help:

Usually, when screwing things in, resistance at the beginning indicates that it's cross-threaded. In this case, it was the opposite; a smooth screw-in would prove to be cross-threaded—tight and skewed after a few turns. On the final, successful attempt, it was tight from the beginning and all the way through, but it did screw up flush with the mount.

And once you've got it going, keep the extension underneath the mount so shavings don't fall on your sensor.

I have a spare unit. I really think about getting yet another lens and thread prolongation to try it over with the image sensor more carefully. I am pretty sure I messed up

Your max zoomed out image without colour boarders yet, does it have image boarders or is it somewhat center?

I don't get colour boarders at zoom, but when I move the image too much (e.g. moving the image all down, at some point gives me boarders)

 

[lrdgreen89]

I have a spare unit. I really think about getting yet another lens and thread prolongation to try it over with the image sensor more carefully. I am pretty sure I messed up

Your max zoomed out image without colour boarders yet, does it have image boarders or is it somewhat center?

I don't get colour boarders at zoom, but when I move the image too much (e.g. moving the image all down, at some point gives me boarders)

At my maxed zoomed-out image without color bars, it can't show opposite borders (top and bottom, or left and right). It's getting most of the image, but it's losing too much for satisfactory archival purposes.

 

[todd95008]

Even before I attempted my first scan, it was my impression that the take-up reel had too much pull, so I took the take-up section apart and removed a washer to increase the clutch slippage. It occurred to me that these films are a lot more fragile than they were 60 plus years ago.

John

New here so excuse me if my post is different for other versions (I have an A version per 0Dan0's list).

Popped the back off and removed the motor assy.
Removed the snap ring and the second large gear.
Removed the 4 screws and black retaining cap (I guess you could call it ) and the 3 washers (2 spring washers and a flat sandwiched between them).
Four tiny screws remove the metal plate and the whole take up reel arm comes out.

The 3 washers mentioned above have nothing to do with acting like a clutch or slippage mechanism !
With the arm completely removed, I still have the same amount of force required to make this slip.
I took apart the take up reel arm and inside there are 2 notched gears off the 2 shafts that are connected by a notched drive belt.
If you apply enough force the belt will slip and jump over a notch or two.
This is the ONLY slip mechanism in the entire take up reel arm/motor assy.

Also, the motor itself is very stiff and does not move.
After seeing this I would not recommend anyone manually move the take up reel as this will eventually damage the drive belt !

Other units may vary, but from pictures I have seen I don't think so ?

My fist scan had lots of frame jumping similar to the above video but since most my reels are 400 feet, there is no way I'm using the "let it collect in a bin" method.

Next I'll try removing the left spring from the hold down door.
I wish I had measured the voltage to the take up reel motor during a scan before putting the back cover back on.
Perhaps I can lower the voltage to the motor a bit and slow down the take up reel ?

Have not put in any modified firmware yet...