Martin Walters wrote:
>Moose: I take your point about metering highlights; however, with an
>OM-2, I would need to use manual mode be able to meter on some discrete
>section of snow to get a "spot" reading, but there's so much snow.
>Otherwise, I would have to use exposure compensation, which is going to
>be trial and error, though in future I would start at + one-half stop.
>
Not likely enough with so much of the white stuff dominating the picture.
>I must admit that deliberately over-exposing snow is counter-intuitive
>for me. I have always been more concerned about bright snow "tricking"
>or overpowering the meter and underexposing all the darker/shadow
>areas. I suspect that when there is bright snow (ie, with sun) there is
>no need to over-expose.
>
Not really. I know it's counter-intuitive, but as Richard already
posted, that's how it works. I hoped the examples from the OM-4 manual
would help.
The way it works is that the meter has to make an assumption about what
it is metering. That's where the whole 18% gray card thing came from.
Some folks, probably at Kodak, a long time ago, measured lots of
different typical photo subjects and determined that the overall average
brightness of the scenes was 18% gray. So the meter assumes that average
brightness. Fine most of the time. But when faced with a subject with an
average brightness much higher, it doesn't know the difference unless
you tell it with the exposure compensation dial or highlight spot
metering. So it happily goes ahead and underexposes the very bright
scene to make it 18% gray average on the film. Then you get something
with a whole range of subtle grays in the majority of the scene, and no
shadow detail. This is technically correct in a way, but not what we
recall seeing and want reproduced in our photograph. The solution is to
increase the exposure enough to place those few parts of the image that
are actually 18% reflective at the correct density on the film. This is
going to lose a lot of highlight detail, but we didn't see it anyway
when looking at the scene, because of the way the human vision system
works, so we don't miss it.
Other solutions are: Use an incident meter to measure the incoming light
directly. Meter an 18% gray card in the same light and use that
exposure. Substitute the palm of you hand for the gray card and increase
that exposure reading by 1 stop. Here the same principle I just laid out
applies. The average human palm reflects a stop MORE light than the 18%
gray card, so one must OPEN UP a stop to get back to average.
> In the past when I used my OM-1 and slide film,
>I consistently under-exposed slightly (say, about one-third stop) under
>all conditions of "normal" light, with absolutely no problems.
>
I would guess you didn't have all that much snow in them.
>So, it may be that exposure compensation would depend on the light levels and
>the extent of snow in the picture (or any other white object, I would
>imagine).
>
Generally true, although the usual amount of light colored subject
matter is actually needed to match the average.
>Interestingly, the last batch of photos that I got from Costco
>on CD were slightly over-exposed, compared to their matching print.
>Obviously, they weren't so this time.
>
Those automated scanners are fickle. I was getting fine scans from a
'real' photo shop with their own machine for some time. Then they got
weird. I've just given up and scan them all myself. No surprise that the
prints are more consistent. It's a different system and one that has
been honed over billions of prints. Its handlers are well trained in
maintaining consistent results.
>I thank you for your efforts with the two photos. Interestingly, the
>one with the house was taken with the non-Oly digital camera and has
>already suffered from brightening that I had already done (photo was
>taken in failing evening light).
>
Yes, I knew that; obvious from the different pixel size. I did one of
each to show that the issue is exactly the same with both kinds of camera.
>I have been playing with other photos
>on my computers at home and at work with a variety of software (PS
>Elements, Corel Photopaint and LView). Each has several menu methods for
>adjusting and enhancing photos, though they may not all do it in the
>same way. The photos can certainly take more adjustment than I had done
>on the versions that I posted.
>
It will work better with 16 bit scans.
>Of course, the other "issue" is that I
>have no way of calibrating my monitors, so with both beiing different
>sizes and with different resolutions, it is hard to compare adn evaluate
>absolute changes.
>
Yeah, it's a problem, but you should be able to get them fairly close in
color, contrast and brightness fairly easily with Adobe Gamma or the
other freee sites/tools on the web. Real calibration is expensive.
>Ultimately, I will have to do some print runs to see the effects of any
>adjustments.
>
Since most of my images are seen primarily on computer screens, I don't
see a print as the ultimate arbiter. And remember, that automated
printing machine is thinking in generally the same way as the camera
meter, working to averages. It is 'smarter' than the camera, but still
doesn't know what it is printing. These machines, esp. those at real
mass market places like Costco, are intentionally set to try to make
normal looking images out of highly overexposed neg film. Without that,
the whole cheap disposable camera business wouldn't work. One stop, one
shutter speed, WIDE latitude film, set it for normal exposure in
moderate light and let the latitude and processing machine take care of
everything brighter. If it's dark, use one with a flash. So you can
overexpose those snow scenes like crazy, and it will still try to make
them overall gray. Automated prints aren't such a good way to judge
exposures of scenes as extreme as these.
>It was so much simpler with slides......
>
Yeah, it's still a fine option. But you always have to choose whether to
lose highlights or shadows in such high contrast situations. I prefer to
make that choice at my leisure in scanning/processing, where I can even
use a curve to compress, but retain both ends and increase contrast in
the middle. Others prefer to do it in the moment with less leeway.
Moose
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|