The perforating pins and matrix have been built to an accuracy of 2/1000th of a millimeter over a length of 9mm. The people who made these are retired now and upcoming fine mechanics are not challenged to such a degree anymore because today electronics supervise mechanics and automatically correct mispositioning. Our machine has to be so precise that mispositioning cannot occur in the fist place because back then no electronic supervision was possible.
This reminds me of some stories from working as a cinema technician.
Not only are the pins and matrix machined to sub-micron tolerance, they are custom machined as a matched set.
One of my jobs as a cinema tech was to maintain and repair splicers in the booths that I serviced. I visited each one of my theaters every 30 days, on average. Each time I arrived, I would go to the storage cabinet and pull out all the parts from broken and disassembled splicers that people had taken apart and couldn't put back together again.
A cine film splicer LOOKS like a simple thing but it is a very difficult thing to assemble and calibrate properly.
The pins and matrices are numbered and they only go together one way. Once you know how to do it, match the serial numbers and look for the alignment marks, they go together smoothly but get them out of order and they'll spite you! What's worse, if they don't fit and you force them together, you'll wear away the cutting surfaces and they will never work correctly again.
You have to assemble the two pieces then bolt the unit to the bottom plate but don't tighten. Then, you align the punch plate on the guide rods and bring it down on top of the top punch. (The pins.) Once you get it aligned, bolt and tighten. Check alignment and tighten the bottom plate and test for correct meshing and operation. Once you get that far, you can slide the top punch/pins assembly off the guide rods to install the springs. Put it all back together and attach the cutter blades and the fulcrum/handle.
I used to just collect all the broken and disassembled splicers in a box. When I got home, I'd dump them out on my kitchen table and use whatever parts I had to make as many working splicers as I could. I'd put them into boxes and stow them in my car. When I visited a theater, I'd get a box out of my trunk, deliver a splicer and pick up the parts of the broken ones that were invariably lying on the shelf.
Just fixing splicers that make one cut on one frame at a time is frustrating enough. At the end of the day, I was terribly frustrated.
I can hardly imagine what it would be like to fix a machine that punches film at a rate of 100 feet per minute!