To me the main question is, who will buy this? What is the market demographic?
As this is going to ba a ‘budget’ or ‘kit’ device, serious archives with real budgets probably wouldn’t touch it anyway, so a sprocket drive is probably fine.
Some people who might want this are:
- Home tinkerers
- Broke people
that the existing systems are too expensive - Unintentional archive with low funds
- Small business (scanning people’s home movies)
- Small Film-makers (shooting on S8 or 16 and finishing on PC)
- Film collectors
For the archive such as the one that sparked the Kinograph in th first place, where you have hundreds of reels and are not sure what is on them, a Kinograph would let you quickly digitise that archive, work out what is worth preserving and perhaps use the cpatured footage to raise funds for a more professional scan and restoration of the ‘worthy’ reels.
Speed here will be the key, as will capturing sound if it is on the reels.
For the ‘small business’ speed is also paramount, as it dictates how much scanning you can get through a week, faster = more income, and more competitive pricing.
The small film-maker probably wants quality as their most key feature, but their film will be in ‘pristine’ condition, so an IR pass will not be necessary, and the system could run slower if required.
The tinkerer, it probably doesn’t matter as much as far as speed and quality are concerned, it is more important that it can be modified, modular and made to work with whatever projects they might envision.
Film collectors may want to ‘backup’ their films, or save a film that is going bad due to VS or similar, or just able to enjoy the film without running their ‘valuable’ master.
So something realtime, with a variable light source, that can handle minor damage and has a good image quality and is cheap to make and easy to modify and repair.
All users will want RELIABILITY, capturing film is torturous enough with cleaning, resplicing, fragile film, trying to keep gates clean, adjusting for the myriad of variables, you don’t want a clunky machine to add to the frustration.
My gut-feel for how that could be delivered on a budget would equate to:
- Continuous variable speed drive - 6/18fps/24fps/25fps drive
- Flashed RGB light source to freeze the frame
- RGB light source to allow colour adjustment for faded or unbalanced film
- Slit-scan LED for audio capture
- Gate with no pressure plate
- Different kinograph models for different formats (8, S8 and 9.5 could share, 16 and 35 could perhaps share)
- Sensor and LED module interchangeable for different Kinograph models
- Integrating sphere light-source for scratch minimisation
- Large sprocket drive to handle damaged/missing sprockets OR roller driven
- Hall effect sensor on large sprocket wheel with embedded magnets to trigger on each frame
- Bayer sensor with 3.4 micron or larger ‘pixels’
- Sensor options from 720P and up to 4K suit price brackets - upgradeable
-USB3 or GigE up to 25fps capture
-Windows/Mac/Linux software to capture, adjust trigger delay, speed, LED RGB mix and flash rate. - Schneider Componon (-S) 50mm lens or similar
- Adjustable focus, with preset known stops.
I know that it would be possible to produce a very good, software controlled light source for under $500 that includes the Windows software to handle all of the above, as well as gamma control and LUT creation, and the sensor could start at US$395, the lens is usually <$100, and the magnets and hall effect senor are <$10, so you could have the lighting, software, sensor and optics covered for about $1000 or less, the rest is mechanics, motors and control boards.
The mechanics could potentially be 3D printed, the ‘gate’ should be machined metal, but could potentially be 3D printed and polished.
I would think you could get away with under $2000 if clever, certainly $2500-$3000 to do something great.
The alternative is to try and make a sub $500 device, which means poorer optics, slower capture, smaller pixels/less dynamic range/more noise, less reliability, you could use a $100 sensor, an intermittent gate, a fixed high CRI white LED, you could evn make it hand cranked, but I think it would end up a curiosity, or maybe let some people who only have a handful of home movies capture something, but it would be little better than poiting the projector at a good quality piece of inkjet photo paper and videoing the resultant image.