Latitude +Brightness range + ND Filters

[Dan Riordan]

Hi,

 

I'm trying to get a grasp on how exposure works.

 

I'm shooting a film on 16mm and I want to get some shots with a really shallow depth of field. If i throw in an ND and open up to say a 2.8 i know I can achieve that, but how does this affect my image in terms of latitude and brightness range? Would that mean that highlights will be blown out? Or will the image still retain it's proper exposure that it would have had at the normal f-stop before I added the ND?

 

What will happen to brighter areas of the image that now exceed the most open aperture setting?

 

I think i'm thinking of the f-stops existing in a fixed position on a characteristic curve, where let's say a 5.6 is fixed at the center. But does properly exposing the image at a 2.8 effectively move the 2.8 to the center of the curve? I'm confused.

[Andrew Brinkhaus]

Dan, good question.

 

Try to think about it this way, the average range for "middle" or "normal" exposure is considered 18% gray. This is also what gray cards are calibrated to. If you use the Zone system of "placing" a level of brightness on a scale, 18% gray would be right in the middle.

 

I wouldn't correlate a certain f-stop with a particular level of brightness, but rather think about it like a see-saw, if the image is darker, your neutral gray, or 18% will be higher on an aperture scale, more wide open. On the other hand, if your image is brighter in exposure, your 18% gray will be more on the lower side of the aperture scale. This also can get confusing when introducing different metering modes like incident, and spot metering.

 

For example, lets say you are lighting a scene, and you spot your gray card, and get an f/2.8, well this means in order to expose that gray card at perfect zone 5, or middle gray, you should shoot at a f/2.8. So having a middle gray, you can now interpret your formats range in latitude to determine how far above/under you can expose and still retain detail. Always remember though, that spot metering is simply trying to tell you how to make a certain item 18% gray. If you are trying to shoot a black couch, and spot it at an f/2, and shoot at an f/2, your couch will appear gray and milky. This should always be considered.

 

Your question about ND is VERY dependent on the format being used. Lets say 16mm, a standard 500T stock, we know that we have roughly 9-10 stops, which means 5 stops over, and 5 stops under. (This is largely dependent on the stock, as some are more/less sensitive/top/bottom heavy). If you wanted to shoot the scene at an f/5.6, and pulled out your meter and read a f/5.6, then you know that area of your image is right at proper exposure. The next step would be to then figure out your darkest areas of the image, and your brightest. Lets say you get an incident reading at the window, and it says f/32. Well, if we have 5 stops above our 5.6 to capture detail, then we know that our reading is right at 5 stops overexposed. To keep from looking overly bright, and to not lose detail, I would try to keep the window around no more than 3 stops over. (This always depends on the mood and setting of the shot!) So we want the window to read at an f/11, which means we need to bring the window down 2 stops. Well, an ND .6 gel is a 2 stop reduction, so that would put me right at my desired exposure.

 

Hope the example helped put it into an easier template, sorry about writing you the book. Ah, I'm sure someone will find it useful.

[David Mullen ASC]

ND means "neutral density" -- it just cuts down the amount of light hitting the lens. So the amount of light reaching the lens and film should be the same when, for example, your correct exposure is a T/5.6... and so you shoot at T/5.6 with no filter or T/2.8 with an ND.60 filter.

 

You control the amount of light hitting the film with a combination of the lens aperture and the filter density (along with frame rate, shutter speed, etc.) So you can create the same net amount of exposure by increasing the amount of ND but opening up the lens aperture by an equal amount.

 

The latitude and contrast should be the same. However... there are some differences in look that are created besides depth of field.

 

Some lenses get slightly more flarey when shot wide-open, losing a little contrast. And when large portions of the frame go out of focus, those areas look lower in contrast because your brights and your darks have merged into more out-of-focus midtones. You can see this on a waveform monitor as you throw a picture out of focus -- you still have the same peaks and valleys, but more of the signal is bunched up in the middle when the picture goes out of focus. So this creates the illusion of overall lower contrast, though the in-focus subject has the same contrast as before when shot stopped down a little more.

[Phil Gerke]

Your question about ND is VERY dependent on the format being used. Lets say 16mm, a standard 500T stock, we know that we have roughly 9-10 stops, which means 5 stops over, and 5 stops under. (This is largely dependent on the stock, as some are more/less sensitive/top/bottom heavy). If you wanted to shoot the scene at an f/5.6, and pulled out your meter and read a f/5.6, then you know that area of your image is right at proper exposure. The next step would be to then figure out your darkest areas of the image, and your brightest. Lets say you get an incident reading at the window, and it says f/32. Well, if we have 5 stops above our 5.6 to capture detail, then we know that our reading is right at 5 stops overexposed. To keep from looking overly bright, and to not lose detail, I would try to keep the window around no more than 3 stops over. (This always depends on the mood and setting of the shot!) So we want the window to read at an f/11, which means we need to bring the window down 2 stops. Well, an ND .6 gel is a 2 stop reduction, so that would put me right at my desired exposure.

 

Hope the example helped put it into an easier template, sorry about writing you the book. Ah, I'm sure someone will find it useful.

 

Not to high jack the thread, but couldn't an incident reading at the window be misleading? What if the light is only serving to increase ambiance? I know you are supposing the window to be in the shot, so I guess what I am asking is what is the best way to meter for a window in the shot? Would not a spot reading of what is out side serve as a better means of determining how to expose or ND for detail through the window? An incident will only give a reading of the light coming through the window and not necessarily what is falling on the objects outside. I realize we are probably talking about a very small difference generally, but just for the sake of asking. Its kind of outside the realm of the original question, but it caught my attention.

 

Thanks a lot. Lots of questions from me today.

[Dan Riordan]

Thanks for your help guys. It's helped a lot. I'll be shooting a short film on 16mm soon and wanted to clear some things up. Can't wait to shoot!

[Saul Rodgar]

Not to high jack the thread, but couldn't an incident reading at the window be misleading? What if the light is only serving to increase ambiance? I know you are supposing the window to be in the shot, so I guess what I am asking is what is the best way to meter for a window in the shot? Would not a spot reading of what is out side serve as a better means of determining how to expose or ND for detail through the window? An incident will only give a reading of the light coming through the window and not necessarily what is falling on the objects outside. I realize we are probably talking about a very small difference generally, but just for the sake of asking. Its kind of outside the realm of the original question, but it caught my attention.

 

Thanks a lot. Lots of questions from me today.

 

Yeah, technically you have to be consistent. For the best results, one would have to use a spot meter, metering the sky, folliage or buildings (whatever is outside really) focusing on something that is 18% gray or as close as possible for your reading to be accurate. And then do the same type of reading inside (hopefully at the right angle) measuring the light being reflected off an 18% card or similar object under key light.

 

This since incident metering measures the intensity of the light falling on an object, and one cannot measure the intensity of the light falling on the sky or a cloud. So unless one can be climbing out to measure the light falling on clouds, it is easier to reflective meter the intensity of the light being reflected by an object, in this case a hypothetical 18% gray cloud in the sky (or whatever is outside that is 18% gray) to know exactly how many stops it is over the interior 18% gray card and judge the correct exposure.

Edited March 10, 2008 by saulie rodgar