After my proof-of-concept prototype gave reassuring results, I started a full-scale project. The very first step I took was reverse engineering of the Lubitel camera I had. I measured main optical parameters: distance between optical axes of the lenses, position of the mirror and the ground glass, and the lens-to-film distance.

Then I switched to the Kiipix developing unit and became surprised with two things. First, I realised that we load cartridges and not films into Instax cameras. Second, there is no information on how film sheets are aligned inside the cartridge. I have to admit, that Fuji release few specifications on their Instax film: ISO speed, image size, and number of exposures per pack - only superficial data. In order to estimate cartridge-to-Instax-film distance I did the following. I reloaded an empty cartridge with ten developed exposures and put the cartridge back into the Kiipix unit. I put eight marks on the unit’s flange around the film window, then I measured flange-to-film distance using a vernier caliper. I measured the distance from the frame edge to where the caliper just touches the film. In order to reduce measurement error I repeated measurements for each film sheet and averaged the results. I pulled each sheet in a standard way by operating the unit knob. I did so in order to produce film movement inside the cartridge to replicate actual use.

The results of my measurements are nearly shocking. Among eight marks where I measured the distance, calculated averaged values ranged from 5.515 mm to 6.2125 mm (0.70 mm interval), and standard deviation varied from 0.277 mm to 0.400 mm (0.55 - 0.80 mm interval). It might not be a big deal, but from a book on 35-mm cameras I learned that the film always bends 0.03-0.05 mm towards the lens and that fact is accounted in the designs. It turns out that Instax film bends ten to twenty times more than expected by industry standards. What is the practical significance? I consider it to be enormous. For example, the main lens moves slightly more that one millimetre while changing focusing distance from infinity to minimal (1.3 metres). Instax film sheets placed inside the cartridge remind me of quantum mechanics particles - you expect them to be discreet and tangible things, but you can describe their physical position only through the probability of being in certain place. In other words, we can only guess how the sheets are aligned. For that reason I just averaged my measurements and used the derived distance in my design.

That finding helped me explain the following feature of Instax cameras: they have automatic exposure control, but a small fixed f/12.7 aperture. Why don’t they use other apertures like f/5.6 or f/8, and why don’t they change the aperture? My assumption now is that Fuji engineered the cameras with small aperture to compensate for both fixed focus and film random alignment in the cartridge. From my perspective, that decision leads to dull images - photos have no distinguishable fore- and background planes. Frankly speaking, the photos look flat.

What is the main take-away? Instax stock lenses made of plastic are apparently bad, and do not utilise the full potential of Instax film. But how much improvement can be achieved by switching to better lenses? I’m confident that certain improvements can be achieved, but at some point cartridges become blockers. Therefore I’m not a fan of transplanting top-end lenses to Instax cameras. I find Lubitel lenses to be the best match for Instax film. Firstly, they are good enough to tap into the potential of Instax film, at least as fat as the design of the cartridges allows. Secondly, the first point still applies even to worn lenses with noticeable backlash in the thread. And last but not least, they are very affordable.

This conclusion fortified my willingness to invest effort into the project.