Some guidelines to exposure compensation in reflected metering

[fran__ois_p._weill]

Here are some guidelines to exposure compensation in reflected metering

 

<p>

 

They might be useless to old hands but I hope they�ll be useful for the less experienced participants.

 

<p>

 

Some examples of things to be considered like 18% gray in true life, hence no compensation necessary:

 

<p>

 

Green fields and trees

 

<p>

 

Brown earth

 

<p>

 

Fall foliage

 

<p>

 

Blue sky

 

<p>

 

Sun tanned faces

 

<p>

 

Typical exposure compensation for some subjects in f-stop values

 

<p>

 

Fog: OPEN 1 stop

 

<p>

 

Sand: OPEN 1.5 stops

 

<p>

 

Snow: OPEN 1.5 to 2 stops

 

<p>

 

White flesh tone: OPEN 1 stop

 

<p>

 

Palm of hand: OPEN 1 stop

 

<p>

 

Overcast sky: OPEN 2 stops

 

<p>

 

Blackboard: CLOSE 2 stops

 

<p>

 

Black flesh tones: CLOSE 1.5 to 2 stops

 

<p>

 

Remember in any case things lighter than 18% gray will need to OPEN the aperture (or slow down the speed) to be properly rendered and conversely things darker than 18% gray will need to CLOSE the aperture (or increase speed). Lighter things to be rendered lighter than the 18% gray need to be exposed MORE than the reading and darker things need to be exposed LESS than 18% gray to be rendered darker.

 

<p>

 

Friendly

 

<p>

 

François P. WEILL

[karl_yik1]

Hi Francois

Dont you mean OPEN 1.5 - 20 stops for black skin tones??

[reinier_de_vlaam]

He doesn't.

 

<p>

 

When measuring the skin tone, the camaera will try to make it 18%

grey. As the skin is darker, you need to under-expose and thus close.

 

<p>

 

For kaukasian skin the recommendation I heard was 0.5 stop, but it

will probably vary depending on the person

[Rob F.]

A weathered (not newly finished) asphalt pavement will give a good

18% reading, too, in a pinch. Blue sky is good, if at least 90

degrees away from the sun.

[matthew_pulzer]

Thanks Francois - useful information.

 

<p>

 

As a rule I am generally pretty suspicious of reflected meter

readings and whenever possible use incident readings. My, quite

literally off the 'cuff technique', is to use the back of my hand as

reference - in my case it seems to match my M6's meter pretty

accurately for 18% grey. Providing my hand is catching the same light

as my subject it works very well, even with exposure accuracy-

sensitive films light Velvia. I do the odd calibration with a

handheld meter, to take account of tanning in the summer, but the

variation seems minimal.

 

<p>

 

The only drawback is you end up looking rather eccentric pointing a

camera at your hand, but then M users coud never be accused of

pandering to convention.

[kelly_flanigan1]

When shooting hockey I open up my lens about 1 stop... (ie one more

stop of exposure)... On an manual camera I set my asa to 400 for asa

800 film...

[roger_michel4]

IMHO, reflected light readings, if used properly, are far more

reliable than incident readings. the film, of course, only cares

about the reflected light. by taking a reflectance reading at the

camera, you are measuring what the film will see. interpreting

incident readings requires the photog to make assumptions about tone

and surface reflectance that may or may not be accurate. it is an

INDIRECT way of measuring what the camera will see. as long as you

understand that a reflected reading is giving you correct exposure

for the measured area so as to render on film as middle grey, you

shouldn't have any problem. learn your zones, learn the size of your

meter area, and you can create any value you want on the neg.

indeed, with a true 1 degree spot and a rudimentary knowledge of zone

values, you can know with absolute certainty how your scene will

render on film. incident readings, by contrast, require you to make

lots of assumptions about how your subject will reflect the lite.

having said all that, if you are shooting a typical scene that IS 18%

reflectance, both systems will work about equally well -- and the

incident reading is less likely to get fooled by backliting.

[fran__ois_p._weill]

Matt:

 

<p>

 

That�s an old trick, mainly useful in backlit situations �

 

<p>

 

And a useful one as long as your skin tone approximate the famed 18%

reflectance�

 

<p>

 

I don�t care for the drawback you mentioned, efficiency is what I

care for :)))).

 

<p>

 

Roger:

 

<p>

 

Sorry but the reflective method has two main advantages: much easier

when using tele-lenses far from the subject, much more prone to

creative interpretation� But certainly not accuracy as far as a

scientific approach is used. The �objective� metering is always the

incident one as it averages the light FALLING on the subject and so

is independent from the reflectance of the subject� This is not

merely my opinion but what has been demonstrated long ago. Notice

incident reading is ever used for illuminance (LUX) metering.

 

<p>

 

What is true in your theory is more than often interpretative

reflective metering will end up with a much better shot than the

direct application of incident meter reading�

 

<p>

 

 

Friendly.

 

<p>

 

François P. WEILL

[roger_michel4]

FPW -- i guess we'll have to agree to disagree about this. while the

incident reading certainly yields an objective measurement of the

lite falling on the subject (at least if you can approach the

subject -- as you point out, you often can't approach the scene in a

landscape shot at infinity), to figure out how that incident lite

will bounce off the subject and record on the film absolutely

requires SUBJECTIVE assumptions about the tone and reflectance of

your subject. only a reflected reading will tell you what the film

is actually seeing. all you have to do then is decided how many

zones above or below medium grey you want your subject to appear on

the film. very direct, very objective, very easy.

[phil_the_small]

snore........

[george_b1]

Roger -

 

<p>

 

I'd have to gently disagree with you on the landscape bit. If there are clouds that diffuse the sunlight on what you're trying to shoot,

and your incident meter reading is taken in bright sunlight, you're absolutely correct; however, if the landscape scene is well-lit by

the same sun that is illuminating you, an incident reading will do the trick.