Modding the Kodak Reels 8mm Film Digitizer (Firmware Hack)
- Thread starter Mac84
- Start date
More success! I've been playing with the UnCamHD online config generator that creates a DCTCONFIG.bin. I placed this on the SDCard with the uengmode file to engage engineering mode and received an error "Cfg file error: Check Magic Err!"
I looked at the bin file and noticed it started and ended with Roral008. Looking for this in the RUnCamHD firmware that is just before the path for DCTCONFIG.bin. On the M127 this is ANEKA001.
I simply replaced Roral008 with ANEKA001, added the DCTCONFIG.bin and uengmode file to the SDCARD, booted the machine and I receive Update config success!
Remove those files from the SDCARD and rebooting the machine, I magically have the folowing file on my SCARD.
CA_DUMP.RAW
With the contents:
It seems we can now set White Balance among other parameters! I'll have a play and see what else I can discover is configurable.
I looked at the bin file and noticed it started and ended with Roral008. Looking for this in the RUnCamHD firmware that is just before the path for DCTCONFIG.bin. On the M127 this is ANEKA001.
I simply replaced Roral008 with ANEKA001, added the DCTCONFIG.bin and uengmode file to the SDCARD, booted the machine and I receive Update config success!
Remove those files from the SDCARD and rebooting the machine, I magically have the folowing file on my SCARD.
CA_DUMP.RAW
With the contents:
Code:
Exposure Value=[2];Exposure compensation,0 : -2EV,1:-1EV,2: 0EV,3 : +1EV,4:+2EV
White Balance=[2];Set White BÿMance,0 :Auto,1: Daylight,2:Cloudy,3:Tungsten,4:Fluorescent,
Contrast=[2] ;Set Contrast ,0 :-2,1:-1,2:Normal,3:+1,4:+2,
Color Effect=[1];Set Color Effect,0 : Standard,1:Mono,2:Sepia,3:Cooler,4:Cool,5:Warm,6:W armer,
Saturation=[2];Set Image Staturation,0 :-2,1 :-1,2:Normal,3:+1,4:+2,
{MicroDVR mov 2012/10/09 v0.49}
{LDMICDVR,v1.1S,01/06/2012}It seems we can now set White Balance among other parameters! I'll have a play and see what else I can discover is configurable.
Annoyingly getting it to dump the CA_DUMP.RAW config seems to have been a fluke as I cannot recreate it. If I flatten the config .BIN file completely it will accept it as long as it prefixes and suffixed with ANEKA001. I just need to work out what the 160 remaining values do I think!
Hello OdanO
Thank you very much for the last B-version 6.1. Finally, the exposure flickering is gone and the video has 18 frames per second. I made a version with RGB=512-256-256 and one with RGB=512-256-192 so that I have a suitable white balance for my films. I need the 2nd version for film material with a blue tint. My wife looks not so happy, because I'm now scanning all the footage again.
I have found one bug in version 6.1: The playback of the movies no longer works on the device.
Thank you for your feedback, and I'm very happy that you making your own white balance versions, and that someone it benefiting from the Type B port. I am aware that I've broken playback (all models), that happened so long ago, and I have no idea which extension did it. I'm probably not going to fix that.
Good day to each other,
Impressed by the possibilities to `modify` this scanner, I have been following this thread for quite a while - now I just say `hello`
On my way to do a lens update, I have successfully disassembled my C-Scanner (currently on version 5.8 of odan0) to at least be able to loosen the bonding of the existing lens. After reassembly, the Reels no longer shows anything on the display - probably because I first plugged the corresponding ribbon cable in the wrong way round... :-(
(the top LED lights up and I can scroll through the menu "blindly" - well - that's no use, of course...)
I now ask myself:
- have I worn out the display...
- ... or have I damaged the ‘board electronics’?
@Hawke
Did you ever find a suitable display for your broken one?
Best regards
yes sorry - translated with DeepL...
The original lens is turned quite far out.
With this focal length, the image has to be scropped for a super8 film.
Would it be possible to get closer to the film with the stock lens and turn it in a little more to be able to use the ‘full’ sensor (i.e. with the whole image sensor closer to the film - instead of further away)?
... can't try it myself at the moment because my display doesn't show anything (probably because of my twisted cable - as mentioned in the post #447)...
With this focal length, the image has to be scropped for a super8 film.
Would it be possible to get closer to the film with the stock lens and turn it in a little more to be able to use the ‘full’ sensor (i.e. with the whole image sensor closer to the film - instead of further away)?
... can't try it myself at the moment because my display doesn't show anything (probably because of my twisted cable - as mentioned in the post #447)...
Hi, unfortunately not. I had put the project on hold. I wrote to two Asian dealers, but they couldn't help me with the data on the display either.Good day to each other,
Impressed by the possibilities to `modify` this scanner, I have been following this thread for quite a while - now I just say `hello`
On my way to do a lens update, I have successfully disassembled my C-Scanner (currently on version 5.8 of odan0) to at least be able to loosen the bonding of the existing lens. After reassembly, the Reels no longer shows anything on the display - probably because I first plugged the corresponding ribbon cable in the wrong way round... :-(
(the top LED lights up and I can scroll through the menu "blindly" - well - that's no use, of course...)
I now ask myself:
- have I worn out the display...
- ... or have I damaged the ‘board electronics’?
@Hawke
Did you ever find a suitable display for your broken one?
Best regards
yes sorry - translated with DeepL...
As the device isn't really cheap and I had digitized everything with the original firmware (not really perfect, but better than nothing
), I wasn't prepared to invest that much again. But maybe that will happen.I wasn't every able to focus the original lens when moved too much closer. Also as you move the lens down, it hits the film loading tray. With the 12mm lens it is really not any lower as the the sensor board is higher.
Why did you disassemble the unit to replace the lens? The Lens can be replace without opening the unit, just remove the lens cover from the front. The reason to open up Reels is for adding the serial port mostly for development purposes.
I still have an issue with my Type B with 12mm lens zoom out to the max.
My original Type C lens hack with a the lens slightly closer:
Both unit are set to sharpness of -1.0. Yet the Type B still lacks resolution, and as weird sharpening halos (resulting in black dots in the lights.)
Clearly something in the firmware's resolution hack is not working on B.
My original Type C lens hack with a the lens slightly closer:
Both unit are set to sharpness of -1.0. Yet the Type B still lacks resolution, and as weird sharpening halos (resulting in black dots in the lights.)
Clearly something in the firmware's resolution hack is not working on B.
Last edited:
It looks like the stock FW when sharpness is set all the way up. I was seeing these dots in highlights as well.
-B
I "blew" my screen on B machine by getting the 40 pin ribbon twisted where it connects too the mother board. After a careful but invasive post-mortem with a microscope, I found 5 conductive traces fried on the screen just beyond the connection where the ribbon transitions to the screen itself. I was able to determine that they were directly opposite the heavier duty power traces the supply the illumination LEDs. The affected traces are used to address the screen matrix that generates what the screen is displaying. Like your scenario, after untwisting and reconnecting the flat cable correctly, the machine booted correctly and seemed to operate, but the screen was either black or unintelligible.Hi, unfortunately not. I had put the project on hold. I wrote to two Asian dealers, but they couldn't help me with the data on the display either.
As the device isn't really cheap and I had digitized everything with the original firmware (not really perfect, but better than nothing
Since then, I've tried to identify the exact 5" TFT 40 pin flat cable screen based on resolution and the exact dimensions of the screen with the assumption that it is an off the shelf item. I found one that fit perfectly, but uses a 30 pin cable. That's the best I've been able to come up with.
Somewhere is a screen with the spec we need or can use with some work-arounds. Kodak does not replace the screen if it comes in for a warranty fix. They just replace the whole machine.
We could use some help out there if someone is savvy enough to find a solution.
well - after opening it I also saw how to remove the cover without opening the part...
But as always - opening things up gives a clearer view
- at least if you plug the cables in correctly afterwards... 
- a new part is here - don't tell my wife...)... I was looking for a solution to lower the sensor - in other words, to be able to replace the original sensor mount using a 3d print...
I somehow overlooked your (unsuccessful) efforts in this regard...
If I look at the 40-pin ribbon cable, which is routed from the motherboard to an ‘intermediate board’, the PINS are routed 1 to 1 to a mating connector on the ‘intermediate board’. The LCD/TFT ribbon cable is connected to this mating connector - but the LCD/TFT ribbon cable probably only has 11 wires (which may indicate control with SPI) - in some cases several wires are pulled together/connected and routed on a single wire within the ribbon cable.
You mention that you have found a custom-fit display (except for the 30 PINs...) - do you have more details?
(presumably the whole thing won't fit anyway because of timings/voltage levels... - but who knows...)
... now I'm a bit confused - would it be possible to get a sharp image with the stock lens using the full sensor size, contrary to the previous tests ? ...
I have this cool firmware from odan0 in version 6.1 on it.
I have removed the scropping completely (`fully zoomed out`) in the settings - I have now gradually `underlaid` the film strip so high that only the actual image is now taken from it again - while I `turned out` the lense out until the image is sharp again. So - as I see it, the full sensor resolution would now be used to capture a Super8 image...
... yes, that's just a blob of felt-tip pen... but amazingly sharp - in my opinion...
Sure, so if I lower the sensor and thus the lens, this film fixer flap gets in the way when I load the film...
... maybe build a fold-away sensor holder with a 3d print...
But back again - am I pushing the lens to its limits with this ‘lowering’ so that effects occur somewhere that I have not (yet) recognized?
I have this cool firmware from odan0 in version 6.1 on it.
I have removed the scropping completely (`fully zoomed out`) in the settings - I have now gradually `underlaid` the film strip so high that only the actual image is now taken from it again - while I `turned out` the lense out until the image is sharp again. So - as I see it, the full sensor resolution would now be used to capture a Super8 image...
... yes, that's just a blob of felt-tip pen... but amazingly sharp - in my opinion...
Sure, so if I lower the sensor and thus the lens, this film fixer flap gets in the way when I load the film...
... maybe build a fold-away sensor holder with a 3d print...
But back again - am I pushing the lens to its limits with this ‘lowering’ so that effects occur somewhere that I have not (yet) recognized?
Yes, but keep in mind the live preview of the image before you record is different from the results of the recording. With the live preview there isn’t as much compression or images to video processing going on, so it tends to look sharper.
There is a “webcam” mode of the device that you can enable via a serial connection, although then the issue would be how to reduce the flicker and frame pull from the video stream etc. Plus I think that stream is only at 1080p.
the subject of ‘video compression’ is currently far beyond my understanding - with this test I was merely trying to find out whether a technical (sharp) focus can be achieved with the stock lens if you bring it closer to the film (ok, in my test I just brought the film closer to the lens)...
The fact that my sharp image is then `compressed` is of course a pity - couldn't the sensor/board be taught to simply create thousands of uncompressed jpegs - even each film image several times with different `shutter speeds` -> HDR
In the end, the film is then only created afterwards in a video editing program with my own compression, which I can then select there...
... that would be a completely new software for the scanner...
Translated with DeepL.com (free version)
it is approx. 18mm from the ‘lens housing’ to the film (the lens itself is approx. 1mm in the housing)
The original distance is about 33mm - right? If you consider that the sensor holder is approx. 15mm thick - you might be tempted to install it rotated by 180°...
admittedly - it still needs some help with the Dremmel (retaining cam and ribbon connector + cable gland) - perhaps easier with the 3d printer...
It won't work the way it's shown here, because the lens already wobbles a little bit in the thread - and therefore probably won't produce a nice film due to vibrations - do you add image stabilization to the firmware?
... I'm a bit conflicted about this test - because I've already ordered a new lens from Scorpio...
Last edited: