Working late in the lab again! I'm trying to detect the film sprocket holes for my DIY cine film scanning project and in the process finding out quite a lot about photo-transistors. The plan is to mount one of these into my film transport mechanism to give feedback to the stepper motor. With some great support from Archie over at
PC-control.co.uk I've made a little circuit to test the sensors and measure the high and low voltages they give.
By placing a bit of card or some white super8 film leader next to the sensors I had the circuit tuned quite well (using the salvaged Epson printer proximity sensor) to the point where I thought this to be a viable route... However, tests with actual film strips gave a very curious result... Though seemingly black to the naked eye, movie film seems to have little or no opacity when seen by the phototransistor.
Realising that the Epson photo diode/phototransistor device uses infrared light, I thought I'd take a look an infrared look at some movie film with my miniDV camera in nightshot mode...
Bingo! The cine film is pretty much transparent under IR lighting. There's no way an IR photosensor will detect the sprocket holes when the film base itself is rendered almost completely transparent. I'll be looking at sensors in the visible light range next...
(Also of interest here, notice how the film looks so scratched under IR lighting; some film scanners use just such an IR pass to enable
infrared cleaning of scanned images.)