Incident Metering Skies for Transparency

[timparkin]

I'm getting my head in a tizzy about how to light meter skies with an incident meter.

Various people are suggesting pointing the meter at the sky, splitting the difference between sky and shadow, etc, etc.

All of it sounds airy fairy and, more importantly, it doesn't seem to work.

When I spot meter, I sometimes need a 1, 2, 3 or (on occasion) a 4 stop (usually a combo).

How do I work out which I need based on incident metering?

Tim

 

[Sirius Glass]

Various people are suggesting pointing the meter at the sky, splitting the difference between sky and shadow, etc, etc.

All of it sounds airy fairy and, more importantly, it doesn't seem to work.

The technical term for that is Bull Crap. Aim the incident meter directly from the subject to where the camera will be. Period.

 

[timparkin]

The technical term for that is Bull Crap. Aim the incident meter directly from the subject to where the camera will be. Period.

That's great for the +/- 3 stops (approx) that covers the 'illuminated' subject matter. But what about the sky? Impossible (without a set of heuristics)?

 

[Sirius Glass]

I always keep the sky out of any reflectance metering that I do. The incident meters are designed to be used exactly the way I describe it.

 

[markbarendt]

There's a couple ways to do this.

The easiest is simply to stand in front of the camera eight or 10 feet and point the meter dome straight back at the camera and take the reading. (Technically you don't have to stand in front of the camera you just have to be able to point the meter the right direction.) This typically works great when the meter can see the light source directly, if the light source is behind the meter it's a tougher call.

Another way to do this is to retract the dome then point the dome directly at the light source, this reading will make sure that you have great highlight detail. By itself this method tends to produce and overall image that's too dark. There are a variety of ways to fix that the simplest of which is to then point the meter directly at the camera and take a second reading. This second reading typically provides a great exposure setting for the mid tones, by itself this method let's the highlights blowout. If you take both readings and then average them you will normally find a happy medium, not too light, not too dark. This works regardless of where the light source is in relation to the camera.

 

[timparkin]

I always keep the sky out of any reflectance metering that I do. The incident meters are designed to be used exactly the way I describe it.

So you're saying it can't be done - thanks...

 

[timparkin]

There's a couple ways to do this.

The easiest is simply to stand in front of the camera eight or 10 feet and point the meter dome straight back at the camera and take the reading. (Technically you don't have to stand in front of the camera you just have to be able to point the meter the right direction.) This typically works great when the meter can see the light source directly, if the light source is behind the meter it's a tougher call.

Another way to do this is to retract the dome then point the dome directly at the light source, this reading will make sure that you have great highlight detail. By itself this method tends to produce and overall image that's too dark. There are a variety of ways to fix that the simplest of which is to then point the meter directly at the camera and take a second reading. This second reading typically provides a great exposure setting for the mid tones, by itself this method let's the highlights blowout. If you take both readings and then average them you will normally find a happy medium, not too light, not too dark. This works regardless of where the light source is in relation to the camera.

That doesn't really help for a sky EV reading though - it assumes that your subject is being lit full face by the sun. Yes this affects the exposure a bit, but only about a stop maybe.

The problem is the hot spot on the white cloud at the top of the picture which could be EV17 in the middle of the day.

 

[David Allen]

If you want to render the sky as the most important thing in your image then forget incident metering.

You need to take a reflective meter reading of the sky with a non-spot averaging meter.

You then simply correct as follows:

• For negative film take the meter's recommended exposure and open up three stops or use a shutter speed that is three stops slower than indicated or a combination of both. This will place the sky as the brightest area with detail on your negative. If you wish to render the scene as Contre-jour then take the meter's recommended exposure and use that.
• For transparency film take the meter's recommended exposure and open up two stops or use a shutter speed that is two stops slower than indicated or a combination of both. This will place the sky as the brightest area with detail on your transparency. If you wish to render the scene as Contre-jour then take the meter's recommended exposure and use that.

It is really as simple as that.

Bests,

David.
www.dsallen.de

 

[markbarendt]

The problem is the hot spot on the white cloud at the top of the picture which could be EV17 in the middle of the day.

If you point the incident meter directly at the sun, dome retracted, and take a reading, then use that reading to shoot with you should essentially eliminate that hot spot. That method protects the highlights better than any other method I know of. It should do exactly what you want.

 

[Poisson Du Jour]

I cannot see any benefit in incident metering the sky. But if the sky is an important element of the overall composition e.g. dramatic light behind equally dramatic clouds, spot meter the lightest (but not pure white), the darkest (but not complete black), and weight these readings 2x or 3x, then proceed to the landscape; weight readings here too, to give the meter something to really chew on. Invariably the sky is a couple of stops brighter than the rest of the landscape, and this is the problem so many photographers have, often erroneously giving too much weight to the exposure of the sky at the expense of correctly exposing all other (landscape) elements. As a rule in my work with Velvia, I do not include plain sky (at any time) in my images; an inspidly coloured sky is best given a complete miss when and where it is possible to do so, or at least reduced so that it is not such a distraction.

The use of ND filters and spot metering must be subject to extensive personal experimentation in-situ; there are many, many variations and outcomes.