Backlit situation with incident meter

[Shinnya]

Hi everyone,

This is one of the situations that I cannot figure out how to get a measurement with an incident meter.

Say, a subject is siting in front of a window where there is a lot of details in the interior that I want to keep, not to mention the face of the subject.

How do I go about getting the measurement with an incident meter? Should I be able to do it without a spot meter?

Any suggestion would be appreciated.

Warmly,
Tsuyoshi

 

[Donald Miller]

Yes, you can meter this quite easily with an incident meter. Take an incident meter reading from the low light position using with the incident dome pointed toward the camera position this will be your exposure...always base exposure on shadow values.

Next, if you wish to determine the brightness range, take a reading with the dome pointed toward the camera position with the dome lit by the window light. By subtracting the low value EV from the high value EV and adding this to five you will have the SBR of the scene. From the SBR you can determine the film EI for that exposure (film EI is not a fixed value but rather a fluctuating value based upon the development that the film will recieve) You will also determine the development that the negative should require based upon the SBR.

This is how I would meter this with an incident meter.

 

[Shinnya]

Donald,

Thank you very much or your a quick repsonse. I can understand the frst part, but not quite the second part...

How can I point with dome torward the camera while lit by the light from the window? Am I missing something?

Thanks!

Warmly,
Tsuyoshi

Next, if you wish to determine the brightness range, take a reading with the dome pointed toward the camera position with the dome lit by the window light.

 

[mark]

Point the dome at the window. Pointing it towards the camera will shade the dome and give you the wrong EV. Not sure if this is what Donald meant to say.

Did you want detail outside the window too.

 

[Donald Miller]

Pointing the dome toward the camera lens will give you the proper exposure for the low values of the scene. Allowing the dome to be lit will give you the proper high value reasding...this can be accomplished by placing the dome perpendicular to camera lens axis.

 

[johnnywalker]

I think he means at the camera, but he said the dome is lit by the window in one case and the "shade" in the other.
Donald, where does the five come from?
I understand you can increase or decrease the EI of the film and change the development time to increase or decrease contrast, but the exact mechanism of doing it under the circumstances the original poster describes eludes me. Would you run through a quick example, using a generic film?

 

[Shinnya]

I am just trying to understand the language here.

So, when I try to read for the highlight value, I would place a meter such that one half of the dome is lit by the window light and the other half pointing toward the camera? Am I understanding it right?

This means that the the direction of the dome will be perpendicular to the way camera is facing the subject is the subject is sitting straight from the camera. Is this right?

Warmly,
Tsuyoshi

 

[mark]

The five comes out of the BTZS incident metering method. At least that is the only place I have seen it. I did not understand the explanation, in the btzs book or video, but it works. I see it as an "ours is not to reason why" type of scenario.

Shinnya, I would place the meter so it is fully lit by the light coming in from the window. It seems to me that, if you place it perpendiclar, half of the dome will be in Shadow thus not giving you a true highlight EV.

 

[Donald Miller]

I am just trying to understand the language here.

So, when I try to read for the highlight value, I would place a meter such that one half of the dome is lit by the window light and the other half pointing toward the camera? Am I understanding it right?

This means that the the direction of the dome will be perpendicular to the way camera is facing the subject is the subject is sitting straight from the camera. Is this right?

Warmly,
Tsuyoshi

That is correct.

For someone who is approaching incident metering with the methodology of spot metering, it appears that this will not work...but it will work since the incident dome gives a range of exposure typically related to the inherent luminance of an 18% gray card.

In this example if you base exposure or development based upon facing the dome toward the scene of maximum luminance you will have an erroneously high reading...resulting in underexposure (if you base your exposure on this reading) or in development (if you base your luminance range on that reading).

 

[Donald Miller]

I think he means at the camera, but he said the dome is lit by the window in one case and the "shade" in the other.
Donald, where does the five come from?
I understand you can increase or decrease the EI of the film and change the development time to increase or decrease contrast, but the exact mechanism of doing it under the circumstances the original poster describes eludes me. Would you run through a quick example, using a generic film?

A typical example using Efke PL 100 developed in Pyrocat HD could be as follows.

The subject is backlit and taking an incident meter reading with the dome pointed toward the camera, I record a EV of 6.6.

Taking a second reading with the dome pointed perpendicular to the axis of the camera lens, I record an EV of 8.6

Subtracting 6.6 from 8.6 gives a value of 2 which when added to 5 gives the SBR of 7. This would be a typically normal brightness range.

From my tests of Efke film I have found that the EI of the film at a SBR of 7 is 50. The correct exposure in this case would be F 11 1/2 at 2 Seconds...which when accounting for reciprocity characteristics would be increased to 4 seconds. The proper development time for the film based upon my equipment, procedures, and desired negative density range would be 11 minutes.

Now had I pointed the meter toward the window light, I would have recorded a SBR of 10.4 and had I followed Mark's original recommendation I would have severely under exposed the film by giving an exposure of F 11 1/2 at 1/8 second rather than the appropriate 4 seconds at F 11 1/2.

If I had followed Mark's second recommendation I would have under developed the film because rather than having a SBR of 7, I would have have decided on a SBR of 8.8 (10.4 minus 6.6 equals 3.8 plus 5 equals 8.8) and I would severely underdeveloped the film. I would have developed the film for 7 minutes rather than 11 minutes.

I hope that this gives you some idea of the procedure that I use.

 

[Michel Hardy-Vallée]

Does your incident meter also have a normal reflected-light reading? Often, meters will have dual-use: 40deg averaging reflected and incident light, by sliding a light dome. If so, then meter the face in shadow (put your meter closer to the face), meter the window, and if you have more than four stops of difference, then reduce development time+overexpose about a stop or so. Testings should be made if you can, of course.

 

[mark]

I stand corrected.

Donald, Does your metering assume the OP wants detail behind the subject? As I understand it, the way you are suggesting to meter this scene would allow for detail behind the subject.

 

[Donald Miller]

Mark,
I am unsure what you mean by "behind the subject". If you mean outside the window, yes the film would exhibit exposure in that portion of the scene since that would carry a luminance that is higher than any other aspect of the scene.

This would be to allow for detail to be exhibited in the backlit subject. It would hold the rim lighting and it would show detail in the unlit aspects of the subject.

With the latitude of the film, typically, it would even show detail in the lit portions outside of the window. This exposure and development would allow us to make a determination at the printing of the negative how we want to interpert the exposure. We could show detail outside the window (if we carried sufficient depth of field) or we could allow it to render without detail...in either case our density range would be appropriate for the backlit subject.

 

[Helen B]

The five comes out of the BTZS incident metering method. At least that is the only place I have seen it. I did not understand the explanation, in the btzs book or video, but it works. I see it as an "ours is not to reason why" type of scenario.

It comes from the range of reflectances of real-world surfaces - from about 3% to about 96%, ie five stops. It predates BTZS. Minor White's beautifully concise Zone System Manual includes the method that would later be better known as the BTZS incident method.

Best,
Helen

 

[ilya1963]

Donald,
At what SBR do you start taking a delution of Pyrocat HD in considaration ?
What paper is this all calibrated for?
Just wonder. ILYA

 

[Donald Miller]

Donald,
At what SBR do you start taking a delution of Pyrocat HD in considaration ?
What paper is this all calibrated for?
Just wonder. ILYA

I don't consider dilution of Pyrocat as a means of dealing with SBR. I would rather diminish development time and under more extreme conditions I would much prefer other means such as preflashing of film for dealing with these extreme brightness situations.

I have arrived at my development times based on the papers that I use (Polywarmtone and Nuance) but more importantly for the enlarger and light source that I use. In my case I use a modified condenser light source.

Having said all of that, I have developed with different dilutions of Pyrocat...for instance 1-1-150 for minimal agitation and enhanced edge effects. Also 2-2-100 when I needed to arrive at a higher density range. But in none of these instances was dilution predicated on the SBR.

 

[Jonathan Brewer]

Hi Shinnya

...............'Hi everyone,

This is one of the situations that I cannot figure out how to get a measurement with an incident meter.

Say, a subject is siting in front of a window where there is a lot of details in the interior that I want to keep, not to mention the face of the subject.

How do I go about getting the measurement with an incident meter? Should I be able to do it without a spot meter?'....................

................My suggestion is to consider measuring your scene, like I do it, with as little math as possible, taking readings in f-stops, with your incident meter set to the same shutter speed as you measure illumination coming from wherever it's coming from as it hits the subject matter/sitter, relative to what your lens sees.

Set your shutter to whatever, aim your incident meter's dome at the lens holding the meter directly in front of the sitter, giving you a reading of the illumination within the room as the lens sees the light hitting the subject.

Turn the meter around and point its dome directly at the illumination coming in from the window and take a reading.

Say the reading inside the room is F4 @ 1/60sec, and the reading pointing the meter out the window is F8 @ 1/60sec, obviously a 2 stop difference.

At some point in considering how you're going to expose this scene, I suggest that you have to mentally switch gears and set the math aside, and consider what's going to 'look right'/going to give you the 'look' you want. If the subject is frontally lit by the illumination coming from inside the room, and is lit by a rim/crescent from the illumination coming from the window, you have some choices in deciding the 'look' you want for this scene.

A subject sitting in front of a window is frontally lit by the illumination coming from inside the room, bias toward the above mentioned F4 @ 1/60sec and the subject's face is going to look less like a shadow, and the illumination coming in from the window is going to look hot.

Expose @ F5.6 or F8 @ 1/60sec and the light hitting the subjects face is going to look more like a shadow relative to the light hitting the sujbect from behind(coming from the window).

The numbers are important, but in order to eliminate mistakes, I simply try to keep the shutter speed the same when measuring at different positions and calculate in F-stops, in the above scenario with a 2 stop difference in your measurements, there's no 'correct' exposure, the best exposure is the one you pick to give you the 'look' you want, or where the shot 'looks right' to.

Having said that, things would get quite a bit more interesting if say measuring the light hitting the sujbect from inside the room, versus the illumination coming through the window, and hitting the sujbect from behind turned out to be a difference of 4-5 stops, picking an exposure between F4 and F16(22), since you want detail on the subjects face without the light coming from the window ending up way too hot or blown out becomes more problematic. Or possibly you might expose at F4 for the illumination inside the room, with light coming in from the window at 4-5 stops hotter creating an interesting effect instead of something garish, it's been done.

If there's a 4-5 stop difference in your scenario, you might want to bias between the two extremes, I'm discussing this assuming the use of neg film and what I've been able to do with a film like Illford HP5, YMMV.

I could play around w/EV numbers, add and subtract, but I prefer a simpler way, particularly when I'm tired/been shooting all day.

Good luck, I hope this helps.

 

[johnnywalker]

Thanks for the example Donald, and the meaning of "five" Mark and Helen.

 

[JBrunner]

As a simple aside, often when using an incident meter, I will "flag" the dome with my hand to eliminate light coming from one direction or another, to give an accurrate value for the area I am metering, in regard to establishing ratios.

For example, if I had a steep backlight that would illuminate part of the dome when oriented towards camera, I would flag the dome with my hand, or block the backlight with my body, to get a value for the subject without the backlight. When using an incident meter, always be aware of the biggest variable of all, your rather opaque self. Metering towards the camera, with the backlight striking the dome would give an approximate average of the front and backlight, depending in part, on how much of the backlight is striking the dome. (It is, perhaps, the least useful of the three readings), and metering towards the backlight would give the value of the backlight. In all instances the values would be for 18% grey, for the light used in the reading.. remember that light striking a subject from an angle that is close to or equal the opposite angle to the camera, particularly if it is a point source, will give a higher resultant exposure in the illuminated area than light coming from 90 degrees and forward on the camera side. The exposure you choose with this information in hand, can be biased in any direction you choose, and the negative will reflect that. If, for example, you expose for the low reading you can expect detail in the low values, at the expense of the highlights ( the window, in the scenario you describe), but with less penalty than if you were to expose for the high reading, and lose the low end. (It's a bitch to print what's not there)

You can also expect lower contrast exposing for detail in a heavily backlit subject, if the source of the backlight is in or near frame, because the lens elements can not be flagged, and the contribution the exagerated exposure brings to backscatter, in regard to exposing the low and middle, in the face of high contrast, where you are letting the highlights go.. (Insert reference to fabulously coated lens here)and
(Insert reference to Pyro developers here)

All things being equal, it's a fine line between clever and stupid, in this scenario.

 

[Jonathan Brewer]

Shinnya....................do some tests w/this lighting situation, bracket your exposures in half stops, starting w/your meter reading for the illumination hitting your sitter from inside the room, keep bracketing until you reach the meter reading you get from pointing your meter out the window. Take some accurate notes of your exposures for each particular frame.

Take what you've heard here, and whatever input you've gotten, and do some involved testing, doing it yourself, and seeing the result, is ultimately going to be the best teacher.