Spotmetering a scene

[marvinhello]

Hi folks, can anyone give me an example of using multiple readings from a spotmeter and calculate the best exposure of a scene? I'm doing a training session on lighting & exposure and kind of confused about this subject.

Thanks!

 

[Grug]

Metering will simply tell you the relative exposures of different parts of the frame for a given sensitivity and shutter angle. So if, for example, you're lighting a nighttime scene and wanted your subject to be a stop under, and they're lit to T/4, then you'd set your aperture for T/5.6. If the background behind them was reading at T/2.8, then it would now be two stops under, and you'd have a fairly dark scene.

 

[Michael Carter]

There's more to it than that though, if you're up to some testing. Ansel Adams' Zone System is a great way to find the DR of your camera and keep your shots within that range.

With histograms these days, it's far easier to get there. But if you know the EV values your sensor can hold in shadow detail, it's a big help.

That said, I rarely even use an incident meter any more, with WFMs and calibrated monitors being so handy. Spot metering is pretty handy for film shooting though.

 

[TheDingo]

A spotmeter is handy to determine the contrast range of your subject, so you can figure out what parts of the scene will be over or under-exposed, and whether you need to add or subtract fill light to lower or raise the contrast range before you shoot. A waveform display does essentially the same thing, though you can't really point to a specific area of your scene on a waveform display.

 

[Policar]

Focus on skin tones. Know where you want to place a given character's face at key, how far under/over darker/bright skin tones are what your contrast ratios are, etc.

It's not about averaging, because you can light to change. When you can't (when shooting wide shots outside let's say) it's about figuring out your over/under, maybe placing grads, etc. accordingly.

Apparently AFI teaches a lot about this, but I have to miss the AFI shoot I am meant to be on.

Maybe ask a graduate of their program? Elswitt still spot meters primarily rather than using an incident.

 

[Grug]

There's more to it than that though, if you're up to some testing. Ansel Adams' Zone System is a great way to find the DR of your camera and keep your shots within that range.

More to it? Like what?

 

[j1clark@ucsd.edu]

More to it? Like what?

The basic problem is that the ISO settings on the DSLR may be 'off' from how the meter is calibrated.

I use both spotmeter and incident meter these days when I'm being picky.

The biggest thing with a spotmeter is 'learning' what is a '18% grey', or some other reference to then adjust the indicated exposure.

For example, most people may not have a 18% grey card handy all the time... for a control narrative shoot this is because someone forgot to pack the test chart... doing shooting in the wild... one can be excused...

Anyway... 'fair/caucasian' skin is about 2x 'reflective' than an 18% grey card, that would be about 36%... unless they've gone for the 'deep tan' look... 'mediterranean' skin is perhaps 18%, and dark complexion may be 9%.

Hence metering the scene to get a exposure, which has people in it, would take these values and adjust accordingly... open up 1 stop from the 'fair', use as read for 'mediterranean', and 'close down' one stop for dark complexion.

As for contrast range, pick the area of lowest values, for example I recent did a shoot in a hotel that had dark wood paneling... that was the 'lowest' large area... walls were off white... the highest value... and then determine the number of stops. My spotmeter has a mode which will give 3 types of readings, 1) relative to the 'low' value, 2) relative to the mid value, 3) relative to the high value. So I would see +/- values relative to those reference points.

All of this presupposes that one knows the characteristics of the camera being used...

In my researches, I've found that for digital cameras the ISO values can be some what arbitrarily determined by the manufacturer. While there is a standard, one 'criteria' reads to the effect of 'what looks good to the engineers'...

There is also the issue of gamma of the camera. In the case of film... ISO/ASA was determined by a minimum exposure threshold to produce a .1 log density above Film Base + Fog, developed for a gamma which placed 18% grey at .5 log density.

There is no 'simple' equivalent for determining the ISO of a digital camera.

What this means is that one needs to test with a 18% grey card, meter that one intends to use, and find such things as 'minimum acceptable 'grey'' say in the Zone System, Zone 3, and then find Zone 7.

It is my belief that most DSLRs which do not represent the data in a 'Log' response curve, are directed to 'direct display' of the image, and as such, 'roll off' high values to achieve 'better' representation. So, in reality most cameras only really represent the scene as a max of Zone 8, which is mostly 'saturated'.

In that regard, placing Zone 5, at a specific IRE % value will determine how many stops above and below that value 'useful' scene information will be represented...

I've been placing my GH-1 18% grey card value at 50% IRE. I then get about 2.5-3 stops above and 2.5-3 stops below and so, I determine that the camera 'really' only has about 5-6 stops of dynamic range.

I know that flies in the face of all the claims to the contrary, but that calibration yields for me a reasonable skin tone for 'fair' while still maintaining good representation for darker complexions.

I've just recently received my Blackmagic Pocket Camera, and have only just begun testing it relative to my color charts, using the camera's Log representation, but research seems to indicate for Log representation 18% grey should yield about a 38% IRE level. I do have a few test chart shots, but over the weekend I just set the camera ISO and meter ISO to the same value and the results are 'ok', because the Wife didn't want me futzing with my cameras on her birthday weekend trip...

 

[j1clark@ucsd.edu]

There's more to it than that though, if you're up to some testing. Ansel Adams' Zone System is a great way to find the DR of your camera and keep your shots within that range.

With histograms these days, it's far easier to get there. But if you know the EV values your sensor can hold in shadow detail, it's a big help.

That said, I rarely even use an incident meter any more, with WFMs and calibrated monitors being so handy. Spot metering is pretty handy for film shooting though.

Oddly... I bought an incident meter after realizing how much lighting in cinematography is listed in terms of footcandles and I wanted a device that read out fc/lux directly... and have been using it much more than my spotmeter, which I use just to check contrast ratios... needless to say, I don't have a camera that has a WFM display, nor a monitor with such either...

 

[TheDingo]

What this means is that one needs to test with a 18% grey card, meter that one intends to use, and find such things as 'minimum acceptable 'grey'' say in the Zone System, Zone 3, and then find Zone 7

I've been using a 20 inch Lastolite EZYBalance collapsible 18 percent grey / white target for several years now, which I find to be better than most grey card targets because it uses a sort of rubberized surface that stays matt ( no reflection ) and is easy to keep clean. The 20 inch target is large enough to be easy to meter, yet collapses down to 8 inches in diameter with a nice sturdy bag to keep it in.

Lastolite EZYBalance Collapsible 18% Grey/White Balance Card - 20"

 

[j1clark@ucsd.edu]

I've been using a 20 inch Lastolite EZYBalance collapsible 18 percent grey / white target for several years now, which I find to be better than most grey card targets because it uses a sort of rubberized surface that stays matt ( no reflection ) and is easy to keep clean. The 20 inch target is large enough to be easy to meter, yet collapses down to 8 inches in diameter with a nice sturdy bag to keep it in.

Lastolite EZYBalance Collapsible 18% Grey/White Balance Card - 20"

I've been meaning to get one of those... the Wife has picked up a couple of giveaway 4" little grey patches... but I think I want the bigger one... I've sort of lost track of all my grey cards at the moment. The Wife would use them in her seminars to start people thinking about using a reference... So, for the while all I have is the Spyder Colorchecker.

 

[Policar]

If you can use a gray card, you can use an incident meter. Using a gray card on film is kind of silly, bordering on absurd.

Spot metering a scene is like the analogue equivalent of using false color or something. You know your over/under for a given stock (+5/-5 for film typically) and see where different parts land, with an emphasis on skin tones. It is useful knowing what parts of an exterior typically land around 18% gray, though (brick, dark tree trunks, etc.).

 

[j1clark@ucsd.edu]

If you can use a gray card, you can use an incident meter. Using a gray card on film is kind of silly, bordering on absurd.

Spot metering a scene is like the analogue equivalent of using false color or something. You know your over/under for a given stock (+5/-5 for film typically) and see where different parts land, with an emphasis on skin tones. It is useful knowing what parts of an exterior typically land around 18% gray, though (brick, dark tree trunks, etc.).

While I've not been on any Real(tm) movie sets, from the BTSs it looks to me that DoPs and others use anything and everything during the shoot. I've seen incident meters, typically the popular Sekonic L-758 (which also has a spotmeter option) or it's ancestors/cousins, I've seen grey cards, grey balls, silvered balls, color charts, etc. Older BTSs often glossed over the 'shooting' setup, but occasionally one would see the cinematographer only using a light meter... and naturally only shooting for Film film through out the process to get to projection prints.

I believe this is perhaps due to the use of digital even if the image acquisition is Film film, so as to have an on set calibrated references for color matching and then color grading, especially if CG elements are being added.

For those who do not have a monitor or camera that presents a WF display, metering is about the only realistic thing to do, and was done since the light meter came in to general use.

 

[Policar]

While I've not been on any Real(tm) movie sets, from the BTSs it looks to me that DoPs and others use anything and everything during the shoot. I've seen incident meters, typically the popular Sekonic L-758 (which also has a spotmeter option) or it's ancestors/cousins, I've seen grey cards, grey balls, silvered balls, color charts, etc. Older BTSs often glossed over the 'shooting' setup, but occasionally one would see the cinematographer only using a light meter... and naturally only shooting for Film film through out the process to get to projection prints.

I believe this is perhaps due to the use of digital even if the image acquisition is Film film, so as to have an on set calibrated references for color matching and then color grading, especially if CG elements are being added.

For those who do not have a monitor or camera that presents a WF display, metering is about the only realistic thing to do, and was done since the light meter came in to general use.

Gray cards or color checker charts are usually shot at the beginning of a reel to give a reference for the lab.

The spheres and stuff are for vfx.

DPs will regularly use spot meters and incident meters, many preferring one over the other. Elswitt is a big spot metering guy; incident meters are more common.

The Sekonic is frowned upon by many (though you'll see it from time to time) and the Spectras are generally preferred for their greater accuracy.

So yes, you will see everything... sure...

But why use a spot meter? Not so you can travel with a gray card and get readings you could already get from an incident meter. I meter virtually everything I shoot, spot and incident (and I only shoot as a hobby mostly) but I never carry a gray card nor would I think to.

And yes, I get that the Alexa doesn't have a waveform monitor but it does have false color and ridiculously generous over/under.

 

[Grug]

The basic problem is that the ISO settings on the DSLR may be 'off' from how the meter is calibrated.

I use both spotmeter and incident meter these days when I'm being picky.

The biggest thing with a spotmeter is 'learning' what is a '18% grey', or some other reference to then adjust the indicated exposure.

For example, most people may not have a 18% grey card handy all the time... for a control narrative shoot this is because someone forgot to pack the test chart... doing shooting in the wild... one can be excused...

Anyway... 'fair/caucasian' skin is about 2x 'reflective' than an 18% grey card, that would be about 36%... unless they've gone for the 'deep tan' look... 'mediterranean' skin is perhaps 18%, and dark complexion may be 9%.

Hence metering the scene to get a exposure, which has people in it, would take these values and adjust accordingly... open up 1 stop from the 'fair', use as read for 'mediterranean', and 'close down' one stop for dark complexion.

As for contrast range, pick the area of lowest values, for example I recent did a shoot in a hotel that had dark wood paneling... that was the 'lowest' large area... walls were off white... the highest value... and then determine the number of stops. My spotmeter has a mode which will give 3 types of readings, 1) relative to the 'low' value, 2) relative to the mid value, 3) relative to the high value. So I would see +/- values relative to those reference points.

All of this presupposes that one knows the characteristics of the camera being used...

In my researches, I've found that for digital cameras the ISO values can be some what arbitrarily determined by the manufacturer. While there is a standard, one 'criteria' reads to the effect of 'what looks good to the engineers'...

There is also the issue of gamma of the camera. In the case of film... ISO/ASA was determined by a minimum exposure threshold to produce a .1 log density above Film Base + Fog, developed for a gamma which placed 18% grey at .5 log density.

There is no 'simple' equivalent for determining the ISO of a digital camera.

What this means is that one needs to test with a 18% grey card, meter that one intends to use, and find such things as 'minimum acceptable 'grey'' say in the Zone System, Zone 3, and then find Zone 7.

It is my belief that most DSLRs which do not represent the data in a 'Log' response curve, are directed to 'direct display' of the image, and as such, 'roll off' high values to achieve 'better' representation. So, in reality most cameras only really represent the scene as a max of Zone 8, which is mostly 'saturated'.

In that regard, placing Zone 5, at a specific IRE % value will determine how many stops above and below that value 'useful' scene information will be represented...

I've been placing my GH-1 18% grey card value at 50% IRE. I then get about 2.5-3 stops above and 2.5-3 stops below and so, I determine that the camera 'really' only has about 5-6 stops of dynamic range.

I know that flies in the face of all the claims to the contrary, but that calibration yields for me a reasonable skin tone for 'fair' while still maintaining good representation for darker complexions.

I've just recently received my Blackmagic Pocket Camera, and have only just begun testing it relative to my color charts, using the camera's Log representation, but research seems to indicate for Log representation 18% grey should yield about a 38% IRE level. I do have a few test chart shots, but over the weekend I just set the camera ISO and meter ISO to the same value and the results are 'ok', because the Wife didn't want me futzing with my cameras on her birthday weekend trip...

I fear you're grossly overcomplicating things there.

Confirming the sensitivity rating of the camera is obviously important, but that should be a given BEFORE you go on a shoot (it should never crop up as an issue on set), and at any rate only requires four things:

- A lightmeter
- A grey card
- A waveform monitor
- A basic knowledge of the camera's gamma curve

Metering will tell you the relative exposure values of objects in your frame, that's all it can do.

The rest is up to the DoP as to how wants expose things given those relative values and the latitude of the camera (which again, should be known before stepping on set with it).

 

[hscully]

I've been placing my GH-1 18% grey card value at 50% IRE. I then get about 2.5-3 stops above and 2.5-3 stops below and so, I determine that the camera 'really' only has about 5-6 stops of dynamic range..

Thank you for saying this. It is important to rate the camera and it is also important to know actual values in frame. I have found that I'm usually operating in no more than and 8 stop range, more frequently lower, unless I'm in broad daylight, then it's just a matter of what to let go. I am usually operating in that 6 - 8 stop range consistent with what I'm reading on the meter.

I am pleased that DR has improved on cameras. When I hear people saying that a camera ONLY has a 12 stop range, I'm surprised by that. It seems like an awful lot to me.