> -----Original Message----- > From: Mark Roberts [mailto:[EMAIL PROTECTED] > > > "tom" <[EMAIL PROTECTED]> wrote: > > >I think the flash meter is set up that it is only metering > while the > >flash is firing, which is not very long relative to the ambient > >exposure. > > > >So, the ambient that hits the film as the curtains are opening and > >closing, as well as most of the time that the shutter is completely > >open, is ignored. > > I'm a trifle skeptical here. Shouldn't that allow the camera to > inadvertently overexpose? If it ignores the ambient light before and > after the time the flash is actually firing, it would > overexpose because > it's letting the film get full ambient exposure *in addition to* the > full flash exposure rather than *including* flash exposure.
Maybe there's a calculation in there as well? I see what you're saying, but the scenario where the TTL meter is operating during the entire shutter cycle should lead to systematic underexposure, and I haven't seen that. > When you're > using standard flash without compensation you'd end up overexposing > every frame by 1 stop. But when you're using standard flash you're almost always doing it when the ambient is 2 or more stops under your selected shutter/aperture value. > > Additionally, I think the propagation delays between > various parts of > the camera/flash electronic system would make the timing > very difficult > to do the way you describe. Especially given the differences between > various different flash units. Well, the delays are short enough that the combo can control flash durations down to, what, 1/25,000? It seems to me that in either scenario, the camera would have to do some calculations and it's not a strict "start flash, meter, turn flash off" scenario, and it seems to me that the calculations would be more accurate if the meter were started when the flash were tripped. Another problem with the "whole exposure" scenario (I think) is that there would be an artificial weighting of objects on the side of the frame that the curtains start from. In other words, the metering wouldn't quite be center weighted since the flash would always be turned off by the time the 2nd curtain starts. I guess if you introduce a calculation into TTL metering, you could get pretty accurate exposures either way. Shortening the merering time seems more plausible as it improves the acccuracy of the meter reading, and TTL flash metering is *very* accurate. Of course, this is all speculative bullshit. tv
