A couple of months back I developed a roll of FP4+ in caffenol for the usual 15 minutes.... or so I thought. Unfortunately I managed to get the recipe all out of whack (not enough soda - I realised afterward that it was more like a recipe suitable for stand/semi-stand). The result was a strip of celuloid with very little on it, at first appearing as though it was completelt devoid of anything. It was so badly under developed that even the rebate numbers/lettering had disappeared. After closer examination there were about three frames just about visible. I almost binned it, but then thought, well why not try scanning the tiny bit that could just be discerned. After adjustment and a good clean-up in PSE6 I found I had a couple of more or less useable images. They're still not completely clean. Having to expand the histogram so drastically also exaggerated any tiny blemish, of which there were quite a few (partly down to initially abandoning the film before I'd squeegeed it off so any muck floating about in the air probably stuck and dried on the strip. I guess it goes to show how flexible the hybrid film/digital workflow can be.