How to expose for highlights?

Andy61557

Senior Member
Messages
1,343
Reaction score
0
Location
ZW
Let's say there is a very bright spot (sheet of paper with text on it) in the frame. How do I make sure that it's white, but not washed out? If I zoom in on it and lock exposure, it will be medium gray. I usually try to include some surraunding area to average it out, but this is just as much trial and error as exposure compenstation.

Is there a quick and sure way to make sure that some part of your picture is white, but not washed out? Maybe with some external spot meter set to 'white' mode?

I know about exposure compensation, histogram and bracketing. Anything else? I want my histogram to go down just at the right edge, quick and sure!
 
Andy wrote:
[snip]
I know about exposure compensation, histogram and bracketing.
Anything else? I want my histogram to go down just at the right
edge, quick and sure!
Practice.

There's no silver bullet that'll get it right every time. Just practice your technique. The quick review histogram is way better than a spot meter or high-tech light meters and such. Just take a shot, if the highlights are blinking, dial in negative AEC, reshoot, and repeat until highlights stop blinking. It's at least as fast as taking a spot meter reading on the bright spot, estimating how much you need to compensate up to get it near blown-out but not actually quite there, and then re-shooting.

After a while, you'll be able to take a good guess without even re-shooting.

Petteri
--




Portfolio: [ http://www.seittipaja.fi/index/ ]
Pontification: [ http://www.seittipaja.fi/ ]
 
Spot-meter on the white paper. Since you only have a partial meter, you may have to zoom in closer so that the partial meter becomes small enough to cover only the paper. Instant spot-meter!

Then, take that exposure reading and add about +2 exposure compensation. That assumes it's the lightest thing in the frame, of course.

--
http://www.neonlightsimaging.com/artshow/final.htm

Extrapolation from few solid data points is best left to those with years of training and experience in such things.
 
That's what I do now, kind of.

Wouldn't it be a logical thing on one of a good light meters? You point at something and say 'I want this to be just below white' (assuming the meter has voice recognition). Of cource the meter needs to know your dynamic range to move the exposure from medium gray to white, but this should be easy.

Well, I actually didn't expect a magic solution from anybody. Just was curious what other people do. Looks like everybody is doing the same thing: taking an educated guess, based on their experience.
 
Well, that's where a nice 1-degree spot-meter on a hand-held meter would come in handy.

Spot-meter on the lightest part, then on the darkest part, and compute the difference.

If 6 stops or less, take the average and shoot!

With the 1D and a relatively long lens, you can do this in-camera.

--
http://www.neonlightsimaging.com/artshow/final.htm

Extrapolation from few solid data points is best left to those with years of training and experience in such things.
 
It's not an educated guess. It's based upon fact and knowing how the equipment/sysem work.

Your meter is set to render an exposure that will reproduce the scene brightness as about 12.5 % of the maximum 100% reflectance. That's exactly three stops difference. There's always a little bit of error involved so, to err on the safe side, you'd use 2 1/3 or 2 2/3 stops more exposure. That would place the white paper at a point where it would be white with tonality and not bleached out white.

No guessing involved.

The exposure compensation control allows up to two stops adjustment so that might work for you, also.
That's what I do now, kind of.
Wouldn't it be a logical thing on one of a good light meters? You
point at something and say 'I want this to be just below white'
(assuming the meter has voice recognition). Of cource the meter
needs to know your dynamic range to move the exposure from medium
gray to white, but this should be easy.
Well, I actually didn't expect a magic solution from anybody. Just
was curious what other people do. Looks like everybody is doing the
same thing: taking an educated guess, based on their experience.
 
There is a 'multi-spot metering' on 1D(s) where you can set 3 points: highlight, medium and dark areas through the focus points. Then the camera will calculate the exposure. Do you ever use it? Doesn't it do just what I'm asking about?

I know it's 'off topic' here, on 10D/... forum, but it's still a good thing to know.
 
Actually, you can use up to eight points.

All you REALLY need are the brightest and darkest point, though, assuming that doesn't go outside the dynamic range.

I've never used it, really.

What I have done is to meter the darkest area and made sure it was underexposed by 3 stops. In that case, I was intentionally blowing out highlights.

I have also metered the lightest area and made sure it was overexposed by about 3 stops (a bit under). In that case, I was intentionally preserving highlights and letting the shadows fall whereever they fell.

Note that this technique is really only useful when lighting is staying constant. At least if you're shooting "action" type shots, instead of static subjects that will wait for you to set exposure.
There is a 'multi-spot metering' on 1D(s) where you can set 3
points: highlight, medium and dark areas through the focus points.
Then the camera will calculate the exposure. Do you ever use it?
Doesn't it do just what I'm asking about?
I know it's 'off topic' here, on 10D/... forum, but it's still a
good thing to know.
--
http://www.neonlightsimaging.com/artshow/final.htm

Extrapolation from few solid data points is best left to those with years of training and experience in such things.
 
Since the dynamic range of 10D is about 6.5 stops, 3 stop comensation will make the gray to be white. If only I could compensate for 3 stops!
 
If only I could
compensate for 3 stops!
you can, if you are shooting in manual mode. If you aren't shooting in manual mode, then you're not really getting consistent exposure.

What I mean by that is, let's assume you're shooting a pic of someone reading that white sheet of paper and they are wearing a black shirt. If you aren't on manual, then when you include more shirt than paper, the camera lets in more light. If you include more paper than shirt, the camera lets in less.

If you think about it, once you have gotten the exposure where you want it, it shouldn't change as you admit more or less shirt. I always shoot manual, so when I've nailed the exposure, it doesn't change as I change the composition.

I hope this makes sense, and that I haven't unleashed the flamers.

Bob
 
Andy wrote:
I
always shoot manual, so when I've nailed the exposure, it doesn't
change as I change the composition.

I hope this makes sense, and that I haven't unleashed the flamers.
Hi Bob:

Are you kidding? I doubt there is one person dumb enough to flame your your perfectly valid suggestion/statement (now watch me proven wrong).

Regards,

Joe Kurkjian, Pbase Supporter

http://www.pbase.com/jkurkjia
 
By compensating exactly three stops you'll be more than likely blowing out the paper's whiteness . You want to step back a little and use between 2 and 2 2/3 stops.
(And the dynamic range is a lot more than 6.5 stops.)
Since the dynamic range of 10D is about 6.5 stops, 3 stop
comensation will make the gray to be white. If only I could
compensate for 3 stops!
 
1. A few people have measured the dynamic range of 10D and found it to be about 6.5 stops. How do you know it more than that?

2. If it is more than 6.5 stops and you step 3 stops form the middle, why is it too much? I don't want to get underexposed too much: noise.
 
hi,

through the custom fn set the * button (or shutter) to lock the exposure. you can then use that button to lock the AE on the bright spot, then zoom out, compose and adjust the exposure comp for the rest of the frame. if your using a wide angle it gets tough sometimes to fill the metered area, so a little guess work is required, otherwise zoom in on the spot and meter from there.
 
how do you manually meter?
start out in Tv or Av, then remember the reading, switch to M and enter it?

feivel
 
meter)..you will need to increase by +2 stop since the meter reads everything as 18% gray (it will be underexposed if you're not compensate it). Another way, which I often use when shooting 35mm film, is to use grey card.
Let's say there is a very bright spot (sheet of paper with text on
it) in the frame. How do I make sure that it's white, but not
washed out? If I zoom in on it and lock exposure, it will be medium
gray. I usually try to include some surraunding area to average it
out, but this is just as much trial and error as exposure
compenstation.
Is there a quick and sure way to make sure that some part of your
picture is white, but not washed out? Maybe with some external spot
meter set to 'white' mode?
I know about exposure compensation, histogram and bracketing.
Anything else? I want my histogram to go down just at the right
edge, quick and sure!
 
to eliminate one area or another. So multi-spot will average from Hightlight, shadow, and mid tone to get as close to the correct exposure as possible.

What kind of subject do you shoot? YOu can use 10D , as other poster here mention, to actually take exposure from 3 different area and calculate the average. Of course, you need to set 10D in M mode.

Or you can use handheld meter , if you have one, to do the same thing. Some handheld meters have average function to do the calculation for you....
There is a 'multi-spot metering' on 1D(s) where you can set 3
points: highlight, medium and dark areas through the focus points.
Then the camera will calculate the exposure. Do you ever use it?
Doesn't it do just what I'm asking about?
I know it's 'off topic' here, on 10D/... forum, but it's still a
good thing to know.
 
Forget about dynamic range. It doesn't impact your original question about exposing for the highlights. Turn that piece of paper over and jot down the following....

Unless you have light sources in your scen, the most any object can reflect is 100% of the light hitting it. And that's in theory because you'd have to have a piece of paper painted with magnesium oxide to come close to the 100% point. So write down 100%. Below that write 50% (that is a change of one stop). Then 25%, 12.5%, 6.25%, 3.12% and so on. You don't have to go too far to see you'll never hit zero.

The meter in your camera wants to render the object you point it at as a tone that is at or about the 12.5 point. So if you aim the meter at your paper, open up three stops, the paper will more than likely be very close to the upper limit of your valid range. You probably don't want that so you open two (or just a nudge over two) to get your highlights placed almost at the limit but still in an area that has tonal separation.

Forget middle grey...it is a dumb term and tends to confuse more often than not.
1. A few people have measured the dynamic range of 10D and found it
to be about 6.5 stops. How do you know it more than that?
2. If it is more than 6.5 stops and you step 3 stops form the
middle, why is it too much? I don't want to get underexposed too
much: noise.
 
I guess I have to do more thinking and reading, and maybe experimenting. I'm glad to have you all on this forum.
Thanks for all input.
 
how do you manually meter?
start out in Tv or Av, then remember the reading, switch to M and
enter it?

feivel
When you're in manual mode the exposure level shows up in the viewfinder on the same meter that would show exposure compensation in one of the auto modes. Center this (by adjusting your aperture or exposure length), and you have what the camera thinks is the best exposure. Go from there.

(At least, I assume that's what's happening. I haven't used manual mode much yet, but that seems to be how it works.)
 

Keyboard shortcuts

Back
Top