This is a response to an off-list question from Mike Wilson. However, I thought it addressed a few undeveloped lines of reasoning from my earlier posts in the thread, and raised some questions that I can't answer myself. In which case I hope Mike doesn't mind that I post my response to him, to the list as well. Thanks for the question, Mike. -------------------------------------------------------------------------
Mike, In truth I don't know how a digi sensor would handle preflashing, especially overflashing. I think I got a bit muddled there saying that it would clip into the blacks. The nature of digital is that there's no result until it's adjusted for display or printing, in which case everything would get moved out to the black and white limits, and all you'd be doing by fogging the blacks is robbing yourself of bit-depth. OTOH preflashing, if carefully controlled to avoid exceeding the inertia point (the threshhold between no record of exposure and the smallest trace of dark tones being captured) should in theory work just the same as it does with film. I also suspect that preflashing would actually need to be simultaneous to the main exposure, because after each capture the sensor would fall back to its unexposed state. It wouldn't be possible to accumulate any latency of fogging exposure for the next exposure to piggyback upon, unless it was within a single capture. I also don't know if digi-sensors are able to accumulate intermittent exposures. The exposure may need to be all in one dose. Ask an owner :-) Multi-layering may work, but begins to negate the appeal of getting a good quick and easy alternative to using a film scanner. For all that, it may not be necesary to use any contrast control in digital copying, especially if shooting to RAW files which capture 12 bits per colour (on the *istD) rather than 8 bits per colour that JPG or TIF files capture ( I don't think the *istD writes 48bit TIF, don't know really). How many stops is 12 bits equal to? From memory a slide has a 5 stop straight line. Pad that out a bit for the toe and shoulder so let's say 7 stops of image. But remember that it's recorded at a gamma of 1.8, or 1.8 times the contrast of real life, that makes it a density range of 12.6 stops. (I wasn't sure if my reasoning was correct here, or whether gamma 1.8 would only add just under 1 extra stop, so I metered some pro 120 tranny on a lightbox and got close to 10 stops range. In that case 12.6 stops for consumer slide film seems reasonable) The big question is:- Can a digi sensor hold over 12 stops in a single capture? The answer to that will be the opposite answer to:- Is contrast control required to copy a slide with a DSLR? Tell me if you know, please. regards, Anthony Farr ----- Original Message ----- From: "mike wilson" <[EMAIL PROTECTED]> To: "Anthony Farr" <[EMAIL PROTECTED]> Sent: Thursday, April 22, 2004 3:51 AM Subject: Re: OT - Why preflash? (was:Re: DSLR slide duplicator) (snip) > > I understand about preflashing - done it a few times myself. I am just > curious as to why (or if) it is neccessary on a DSLR. I would have > thought that there are other techniques (for example, multiple exposure > overlay) that could be used. I am also unsure of the similarity to film > in this respect of the sensor. > > mike > >
