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

[Rob_A]

A small addendum to ThePhage's splendid User's Guide to 0dan0's Enhanced Firmware...
In the FrameRate section I read "8mm film is 16 fps; Super8 film is 18 or 24 fps".
While it is true that the frame rate for standard-8mm cameras was 16fps in the days of spring-driven ("clockwork") motors, with the arrival of battery-power, circa 1961, most manufacturers adopted 18fps as standard. Thus when scanning standard-8mm film the appropriate setting depends on the camera that was used; older film almost certainly 16fps, post-1961 possibly 18fps.

 

[0dan0]

I've been collecting my thoughts for some comments and queries about the updated firmware and use of the Kodak Reels in general, but this discussion of alternative lenses prompts me to jump in now with a specific question that seems relevant. 0dan0 unhappily reports that "16mm AZ-231622S-MAC hasn't worked, I couldn't get a focus at this short distance". Might it not be possible to achieve focus at a closer distance by adding a supplementary lens? Back in the sixties I used a 1 dioptre "close-up" lens on my Sankyo 8mm camera when shooting animation. Similarly a 2 dioptre supplementary on my 35mm Voigtlander enabled me to copy prints etc. at a shorter distance than the standard (fixed) lens allowed. Granted that we're now considering a MUCH shorter distance than in those cases I suppose that in principle it might be possible to come up with a supplementary that would appropriately reduce the focusing range of the 16mm lens. Others with more knowledge of optics will be better able to calculate the required spec for such a lens, but it seems to me that it might be a feasible solution. Obviously any additional glassware potentially degrades the achievable quality by some margin, but maybe not enough to offset the advantage of an optimised field of view?

This thread has been an invaluable resource while I've been getting to grips with my "Kodak" Reels and I have to express my appreciation of all the contributors, particularly those responsible for the hugely impressive firmware and hardware enhancements. My scanner is as yet unmodified, until I have completed an initial digitisation of my family's 8mm archive (1958-82) but I hope that in due course I'll be able to take advantage of all that voluntary development work.

I have tried diopters. That might work if the focus was almost in, it wasn't close with the above lens. I expected the sensor needs to be mounted higher. In basic optical terms it makes sense, the longer 16mm lens, requires more distance from the front element to the sensor. Moving too much higher is both not practical, and could defeat the goal of greater magnification. A lot of work for a film format that doesn't resolve that much anyway.

 

[ThePhage]

A small addendum to ThePhage's splendid User's Guide to 0dan0's Enhanced Firmware...
In the FrameRate section I read "8mm film is 16 fps; Super8 film is 18 or 24 fps".
While it is true that the frame rate for standard-8mm cameras was 16fps in the days of spring-driven ("clockwork") motors, with the arrival of battery-power, circa 1961, most manufacturers adopted 18fps as standard. Thus when scanning standard-8mm film the appropriate setting depends on the camera that was used; older film almost certainly 16fps, post-1961 possibly 18fps.

Thanks for the clarification and history, Rob_A. I've updated the User Guide appropriately.

 

[ThePhage]

With a Model C on v7.1.1, I started scanning a 7" reel of Super8 film yesterday. Started off going well for about 7 minutes of resulting footage (with no issues on the resulting file). And then I had to stop the capture to re-frame the image. I restarted the capture and it automatically ended after about 17 minutes of resulting footage (the MP4 file is 4.2GB in size). When it ended capture, the scanner's on-screen display'gave me the typical poup message about closing or playing the file. But the scanner seemed to be frozen because it wouldnt' respond to button pushes. I was able to power off the device and the most recent captured file was okay. I am curious if the capture ended because of a 4GB filesize limit. On previous scans of 7" reels (about 9 months ago), my files never exceeded 4GB, even with Mac84's increased bitrate firmware.

Also, I'm sad to report that the entire 17 minute long video had some jitter on the bottom of the image (where I had only previously seen it on the top, when it would occasionally manifest - but that was on a different scanner, with Mac84s firmware). Here's a sample from it:

Still not quite sure what triggers that from happening sometimes, but not other times on the same footage.

 

[sheider]

With a Model C on v7.1.1, I started scanning a 7" reel of Super8 film yesterday. Started off going well for about 7 minutes of resulting footage (with no issues on the resulting file). And then I had to stop the capture to re-frame the image. I restarted the capture and it automatically ended after about 17 minutes of resulting footage (the MP4 file is 4.2GB in size). When it ended capture, the scanner's on-screen display'gave me the typical poup message about closing or playing the file. But the scanner seemed to be frozen because it wouldnt' respond to button pushes. I was able to power off the device and the most recent captured file was okay. I am curious if the capture ended because of a 4GB filesize limit. On previous scans of 7" reels (about 9 months ago), my files never exceeded 4GB, even with Mac84's increased bitrate firmware.

Also, I'm sad to report that the entire 17 minute long video had some jitter on the bottom of the image (where I had only previously seen it on the top, when it would occasionally manifest - but that was on a different scanner, with Mac84s firmware). Here's a sample from it:

Still not quite sure what triggers that from happening sometimes, but not other times on the same footage.

@ThePhage Sorry to hear about your issues scanning 7" reels. I only have 3" reels to scan, so I don't have any 4 GB file size concerns. However, I also had the device "freeze" during a 3" reel scan once... I can live with that, because it hasn't done it since. Regarding the random jitter issues, I can think of a few possible causes:
1. Dirt/gunk accumulating in the gate path during a long scan
2. Not using all of the "rollers" on the right side of the scanner bed
3. Film not under all 4 tabs in the scanner bed
4. Hardware glitch such as a motor hiccup
5. Perforation tear or bad splice on the film
6. Takeup (drive) reel hub not fully seated on the drive spoke
7. Friction: Film stuck together or rubbing against the sides of its reel

Good luck, and don't tolerate faulty hardware if your unit is still under warranty!

 

[0dan0]

4GBytes is a file size limit. At 35mbs that around a 15minutes for an 18fps reel. The software will crash at 4GB, but the file safely closes. Reboot is required. This would practically never happen with stock firmware at 8mbs, so there is no file chaptering implemented.

The glitch is a long term issue, way less on v7.x than v6.x. maybe <5% of captures. It happens on stock firmware, happens a tad more with the 12mm lens. It is just bad hardware/firmware design. I wish I could detect it.

 

[TheElk]

With a Model C on v7.1.1, I started scanning a 7" reel of Super8 film yesterday. Started off going well for about 7 minutes of resulting footage (with no issues on the resulting file). And then I had to stop the capture to re-frame the image. I restarted the capture and it automatically ended after about 17 minutes of resulting footage (the MP4 file is 4.2GB in size). When it ended capture, the scanner's on-screen display'gave me the typical poup message about closing or playing the file. But the scanner seemed to be frozen because it wouldnt' respond to button pushes. I was able to power off the device and the most recent captured file was okay. I am curious if the capture ended because of a 4GB filesize limit. On previous scans of 7" reels (about 9 months ago), my files never exceeded 4GB, even with Mac84's increased bitrate firmware.

Also, I'm sad to report that the entire 17 minute long video had some jitter on the bottom of the image (where I had only previously seen it on the top, when it would occasionally manifest - but that was on a different scanner, with Mac84s firmware). Here's a sample from it:

Still not quite sure what triggers that from happening sometimes, but not other times on the same footage.

I use @0dan0 s V7.11. I recognize sometimes flickering, but this I could resolve in a second scan with one or two steps down or up in framing depending on the location where the flickering occurs. The flicker is not dependend on the software but on the frame located in the scanner.

 

[thjakob]

Now that the days are getting shorter (and the evenings longer), I'm finding time to read here again...

I've now converted my C-scanner to a Scorpio lens.

However, I'm still struggling with the quality of my images...

The preview on the scanner display shows sharp images (the preview is - well, not that big...) – but in the film produced, I think the resolution is much worse.

Firmware: odans0 7.1.1
My settings:
WB: -1
Sharpness: -0.5
Green: 304 (432, 304, 320)
ev: 0
FPS: 18
----------------
The scanner shows:
- ISO 100
- Exposure: 2891

Yes, I know about this “video compression” – but is it really that “bad”?
Does compression also depend on the image content?...
I think that the image quality is quite good in the white opening credits (deepl, are you sure about what you've translated... )...

But when the images start, it somehow becomes “mushy”.
Is that really “grain” – even where there's another bit of fluff – and it seems to me that it's not reproduced very sharply.
(Admittedly, the image may not be really sharply focused...)

I also took a series of shots with sharpness settings ranging from +2 to -2.
In my opinion, the videos with -1.5 or -2.0 look best. But I think Odan0 once said that with the 12 mm lens, a sharpness setting of -0.5 (or 0) would be okay...


Am I doing something wrong?

Thanks & best regards,
Thomas, who is still impressed by what has been happening here over the last few months.

 

[0dan0]

Compression is very much image dependent. Yes you typically resolve more details from the fluff and the scratches than the image itself from 8mm film. At at 35 mbs the quality for noisy source is good, not great. More of a factor is 8mm film itself is not high resolution. The main reason the stock lens is good enough for most users. However can be worth revisiting the information gained in a before and after comparison

(it has gotten better since v6.5 used to make this video.)

I use sharpen at -1.0 mostly with the new lens.

 

[thjakob]

Compression is very much image dependent. Yes you typically resolve more details from the fluff and the scratches than the image itself from 8mm film. At at 35 mbs the quality for noisy source is good, not great. More of a factor is 8mm film itself is not high resolution. The main reason the stock lens is good enough for most users. However can be worth revisiting the information gained in a before and after comparison

(it has gotten better since v6.5 used to make this video.)

I use sharpen at -1.0 mostly with the new lens.

Thank you for your feedback.
Yes, I am familiar with this comparison...
On the other hand, I interpret your answer to mean that I am probably not doing anything wrong.
My concerns about the quality (specifically the difference in quality between the “preview” and the “final product” – I am definitely not complaining about the improved quality of the streams with the “hacked” firmware versions ) are therefore probably due to the video compression. Well, is it possible to obtain “uncompressed” video material from this part?
(Yes, of course, then there's the issue of file sizes again – but...)

Best regards,
Thomas