Modding the Kodak Reels 8mm Film Digitizer (Firmware Hack)

TheElk said:

What Blocksize did you use ? See my answer on post ( #859 ). May be try to put a small file on the card when attached to your PC.

Click to expand...

What you all knew already...

Based on TheElk's suggestion, I formatted a few cards with Fat32 32kb blocks (only possible in Windows with additional tools such as Rufus – Windows does not allow this formatting > 32GB with 32kb block (although it is actually valid according to the FAT standard definition))...
... and what can I say : the scanner scans and scans

maximum test: 64 GB of nonsense written to the 128 GB card – and then continued scanning with the scanner – so far so good

Belongs to the topic: good to know – but not really necessary...



I'll keep quiet for now that this also works with a 64kb block size...

0dan0 said:

The random stopping is known, and happens more with the sharper lens. It will most often happen when the scene has a lot of texture and details, for which the new lens reveals more. The encoder is crashing with so much information (it is a crappy encoder from ~2011.) user control, a lower sharpness can help. Things I can try for this. 1. At least display the error message, so you can know why it stopped. 2. Allow user controlled bitrate, by setting Qp new limit. 3. Lowering the overall bitrate/quality (not planning on.). The issue for 2., I have run out of buttons in their current usage method. I have tried to stop the crashing, but without source code, not much more I can do.

Click to expand...

thanks for pointing that out (sorry – I had overlooked that information before).
Well, we don't want the images to be any less sharp – what's strange is that these previously “broken” areas then slip through on a second attempt... memory leak – or something like that...

It's fine by me – no feature request for me.

I've taken the opportunity to revise the User Guide to 0Dan0's Custom Firmware to include some Known Issues. Will try to keep these updated as progress is made. I've also revised the Version Release notes section.

Also, @eyeidea, while I think it would be great for someone to put together a summary of lens modification efforts, that's one topic I've decided to not add in. For anyone who would like to do that, that would be most welcome and I'll gladly cross reference that information in the User Guide I'm maintaining.

thjakob said:

thanks for pointing that out (sorry – I had overlooked that information before).
Well, we don't want the images to be any less sharp – what's strange is that these previously “broken” areas then slip through on a second attempt... memory leak – or something like that...

It's fine by me – no feature request for me.

Click to expand...

Qp - compression quality - will ramp up over time. Often a tough segment will be at a low Qp (high quality) at the time of the encoder crash. The next capture for the same frames uses a higher Qp, almost always solving the issue.

I have installed Version A7.1.1.
The colors are sensational and no fame flicker occurs anymore. Thank you for your excellent work!
With this firmware, my Kodak Reels allow only to access the Record and the Rewind function. On all the other commands like view my recordings etc., the machine freezes. in addition, I guess that the limit of 16700 frames per recording is due to the limited file size of FAT? My USB- Stick with NTFS works fine with large MP4 files. if these items could be fixed, the firmware would be absolutely perfect.

For more frames we will need to lower the quality. This is not an introduced bug, it is a limitation on the Reels system design. It does not support 64-bit MP4s, required for files over 4GB.

As for the other bugs, like broken playback, they are likely due the increased memory needs for processing at a higher resolution. So more quality vs playback, it was a easy choice for me. If playback is important to others, I've asked this before, someone needs to find the latest version when playback was working.

0dan0 said:

For more frames we will need to lower the quality. This is not an introduced bug, it is a limitation on the Reels system design. It does not support 64-bit MP4s, required for files over 4GB.

As for the other bugs, like broken playback, they are likely due the increased memory needs for processing at a higher resolution. So more quality vs playback, it was a easy choice for me. If playback is important to others, I've asked this before, someone needs to find the latest version when playback was working.

Click to expand...

Thank you for the rapid reply. I can live with these conditions and I think I will stay with your firmware. The advantage is worth it.

It might nice for someone to determine which version broke playback, like which version broke 64GB card support.

0dan0 said:

It might nice for someone to determine which version broke playback, like which version broke 64GB card support.

Click to expand...

@0dan0 - playback broke with v4.0:

V3.0 16 Feb 2025
On-device movie playback works with this version.

V4.0 18 Feb 2025
BUG INTRODUCED: On-device movie playback is now broken... Perhaps this is a result of moving 1 of the 2 image buffers lower in memory???

As I mentioned in an earlier post, my 128GB FAT32 card works fine with all fw versions that I've tested: stock, @Mac84, and yours up to the current version.

sheider said:

@0dan0 - playback broke with v4.0:

V3.0 16 Feb 2025
On-device movie playback works with this version.

V4.0 18 Feb 2025
BUG INTRODUCED: On-device movie playback is now broken... Perhaps this is a result of moving 1 of the 2 image buffers lower in memory???

As I mentioned in an earlier post, my 128GB FAT32 card works fine with all fw versions that I've tested: stock, @Mac84, and yours up to the current version.

Click to expand...

Thanks, v3 to v4.

Do we know when exFAT broke?

0dan0 said:

Thanks, v3 to v4.

Do we know when exFAT broke?

Click to expand...

... is it worth it...?
A note in the user manual on the subject of SD cards would probably suffice...

thjakob said:

thanks for pointing that out (sorry – I had overlooked that information before).
Well, we don't want the images to be any less sharp – what's strange is that these previously “broken” areas then slip through on a second attempt... memory leak – or something like that...

It's fine by me – no feature request for me.

Click to expand...

Regarding image sharpness, I just noticed that my previous assessment of the raw material (i.e. the MP4 files directly from the scanner) depends heavily on the MP4 player used...

With the VLC player, the image is much clearer than when using Windows Media Player.

1: Windows Media Player
2: VLC



...as always – all factors can play a role...

thjakob said:

... is it worth it...?
A note in the user manual on the subject of SD cards would probably suffice...

Click to expand...

Done. See link to the guide in my signature.

I've also added in the complete version history for reference on when certain changes took place.

ThePhage said:

Done. See link to the guide in my signature.

I've also added in the complete version history for reference on when certain changes took place.

Click to expand...

May be you should make a note, that
"Setting will be remembered after unit shutdown" for the 16, 16, 24 fps setting will not be working all the time. I recognized, that often 18 fps are on after reboot. I remembered this, because I in most cases use 24 fps.

The same for the "green tint" . After powering on, often the green setting ( the middle value in "528, 320, 256" here 320, changes to an other value nearby.

Is there any trick to keeping the film in the tracks as it being process?

VoodooKi said:

Is there any trick to keeping the film in the tracks as it being process?

Click to expand...

Check that you have the film fed under the 4 black tabs in the scan path. Depending on how/where the film is coming out, you may need to adjust the tension on the metal press plates by slightly screwing or unscrewing them (the plates are spring loaded). You may also consider not using the take-up reel, but let the scanned film spill out into a large bin / container (avoiding the need for it to wrap around the grey rollers/posts).

The only film I've had that really wanted to pop out was the film that had warped / cupped which causes also causes it to not focus well across the entire image. If the film is nice and flat, it should be able to stay in during the scan.

thjakob said:

Regarding image sharpness, I just noticed that my previous assessment of the raw material (i.e. the MP4 files directly from the scanner) depends heavily on the MP4 player used...

With the VLC player, the image is much clearer than when using Windows Media Player.

1: Windows Media Player
2: VLC

View attachment 24569

...as always – all factors can play a role...

Click to expand...

Some older players are not acknowledging the full range metadata, that is not sharpness, rather an increase in contrast (also a clipped dynamic range.)

Hi all, I've just mounted the newly arrived 12mm lens.
I live in Hong Kong so I've tried to find the lens around here (China). After a couple of fails, I've finally got this one working perfectly through Alibaba:

2/3" F2.8 F5.6 F8.0 10mp FL 12mm 4k M12 Macro Board Lens​

They will sell a 'sample piece' (there is an option in the price section). They took almost 3 weeks to ship, then the good arrived in 2 days to me in HK.
They have different models, I got the 2.8f, but probably I should have bought the 5.6 for a bit more depth of field, but it is what it is...

I've used both a new mount (I've bought a bunch of different lengths from Aliexpress here) and the 7mm extender here. with the 2.8f you need to try different combinations. At the end I was tired of try combinations and I stood with this one:

The lock ring was from the same shop in Aliexpress where I got the extender, here.

To remove the lens cover, gently but firmly push up the bottom and the sides (there are two plastic latches at the bottom and one on each side, and it will rollup.
To detach the sensor from the flat cable, the black segment in the connector must to be flipped up, as shown in the picture (I thought it was one of those that opens by sliding but it was instead the flip-up one):

I've opened the whole scanner to be able to cut the space for the cable when the sensor is at the new position:

And used the wonderful 3D printed holders from @0dan0 (can't thank you enough!)
Last I've focussed, tightened the lock ring and cut the slots in the original cover to put it back: having the camera so exposed made me feel uncomfortable...(if you want to do the same, you will have to rotate the cover upside-down, with the flat side on top and the sloped one at the bottom. In order to do so, you need to trim a bit the latches that keep it in place, because they are not symmetrical):

The new lens is not too long and there is plenty of space between the tip of the lens and the opened loading bay door.
The overall process is not difficult, but some parts will need some repetition (for example for me was the right combination of mount+extender).
Hope this helps with the lens replacement.
Ciao, Federico

I love how you have restored the camera cover. You should add some cover for the circuit board, as there is a small amount of light that can leak through the PCB. In very dark images, with a lot of sensor gain, you can see circuit outlines, which is a tad bizarre. I just did my third unit, with the same F2.8 Lens.



With a piece of black masking tape over the circuit board.

This is a newer version of the 3D print. The vertical attachment to join the two halves was a little fragile, so I made a newer sturdier design that supports top half siding in from the front. Shared: https://makerworld.com/en/models/2000850-reels-scanner-lens-offset



I agree that F5.6 might be a great option, but I don't believe you need a new lens of this, just 3D print an iris for the front of lens.