Jim Maroney wrote:
> fcc wrote:
> "I use evaluative mode and dial in +1 or +1.5 exposure compensation"
> Given the "evaluative" nature of this metering mode is exposure compensation
> effective?
Sure--it just takes the base meter reading and shifts your exposure by whatever
amount of compensation you apply. You could take a meter reading, set the camera
to manual, and then set the exposure yourself with whatever shift you wanted,
and get the same result. Although modern meters are more sophisticated than my
old Gossen Lunasix, they still do the same thing--read the the light of a scene
and recommend settings based on some pre-selected calibration point--in almost
all cases that is 18% reflectance such as you get from grass, light-toned skin,
medium-green leaves, etc. When faced with something more reflective, like snow
or beach sand, the photographer must compensate for the difference because
there's no way for the meter to know it's "seeing" snow instead of a lawn and
will underexpose the scene. It really doesn't matter what mode you're in or what
kind of meter you're using; the only exceptions would be if you were to
spot-meter some middle-toned subject in an otherwise overly-bright or
overly-dark scene, or if you were to use an incident meter.
"Evaluative" is basically a marketing term, like "matrix" (Nikon) or "honeycomb"
(Minolta). All it's telling you is that the meter is using some sort of
relatively complex algorithm to read the entire scene instead of a more simple
scheme such as center-weighted averaging, partial-area, or spot. Dialing in some
compensation isn't going to confuse the meter, since it isn't really thinking.
:-)
Nonetheless, exposure compensation for most of us is largely a "feel" kind of
thing (although I suppose you could find a way to measure the reflectance of the
target and calculate mathematically what the compensation should be--but I
suspect it would be less time-consuming to just carry a gray card), so it's
usually a good idea to bracket a snow exposure by varying the amount of
compensation. You could use autobracketing, but since you know that the normal
exposure is going to underexpose the scene, autobracketing without shifting the
exposure up to about +1 or +1.5 will just waste two frames, since the normal and
"-" exposures will be underexposed. I usually use +1 as a starting point and add
more if it's sunny, less if there are significant darker areas in the scene.
> If so, how does it work compared to center weighted or spot
> metering modes?
I don't use those modes much (although I did use center-weighted when I had my
A-1, of course); I did use all the different ones when I first got my 3, but
after some experience with evaluative I found it so good I don't use the other
modes except in unusually difficult circumstances. The biggest drawback to this
approach is that I sometimes forget I have other metering options, and when I do
use them I often forget to switch back, so that I find I've just exposed half a
roll of film in spot mode when I thought I was in evaluative mode--which can of
course be a problem. :-)
Frankly, IMO unless you're experienced with spot metering, evaluative will give
a better result most of the time since with a spot meter you need to be able to
accurately judge what in the scene is a midtone (by eye), or else be able to
dial in the proper EC for whatever thing you've metered--except that with spot
you need to be more accurate with your compensation since you're working with
just one small part of the scene. Alternatively, you can take multiple spot
readings and then come up with an average reading in your head (although the 3's
multi-spot function does this for you), which again requires more experience and
more thought. The bottom line is that for any situation where you need to adjust
the exposure or to be concerned about weird or uneven lighting, experience and a
good understanding of the principles of exposure are going to be necessary,
whether you're calculating exposure based on a spot reading or dialing in EC
when in evaluative mode.
> I recall having read somewhere that EC was
> unpredictable/ineffective when used in evalutive mode. Am I off base here?
Well, that hasn't been my experience. But since any application of EC is
fundamentally a matter of the photographer's judgement, it's only going to be as
good as that judgement. When in doubt, bracket! I suspect what some folks are
experiencing is that evaluative metering (at least in it's implementation in the
3/1v and perhaps the 7/30) can seem to be unpredictable because it does a good
job of handling tricky lighting and therefore doesn't always need compensation
when you'd think it might (I've had this happen myself). So in that sense it may
be that evaluative mode is less predictable than a more simple system like
center-weighted averaging; with the latter the photographer can, with
experience, know almost automatically when compensation will be required or what
in a particular scene to meter; she learns the characteristics and peculiarities
of the meter. But I suspect the same thing can be done with evaluative metering;
perhaps it's a matter of coming to understand that compensation might be needed
less frequently because the meter does a better job of analysing the entire
scene. Or something like that.
Craig
*
****
*******
***********************************************************
* For list instructions, including unsubscribe, see:
* http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************
