light meter confusion
Posted 15 October 2006 - 07:13 PM
REFLECTED LIGHT METER
Before, with a reflective light meter, I would point the camera at my subject's face, given it being 10 feet away or 100 feet away. The camera would tell me how much light was bouncing off of his/her face and I would usually open up a stop (caucasion skin). This worked perfect for me. I always imagined the camera was reading how much light was bouncing off the face and going into the camera.
When taking readings off of a light bulb that would be in the scene, I would just look at that and zoom in, seeing how much much light would travel to the camera and expose the film. I mean, my light meter IS the camera, so it would tell me how much light would hit the film. This is where I got confused.
INCIDENT LIGHT METER
when using the incident light meter, I use the flat disc. I walk up to the subject's face and hold the disc towards the light source. It tells me how many footcandles, and therefore which f/stop. But, here is the catch: I walk 10 feet away...or I walk 100 feet away, and I set the f/stop to what the light meter was reading. What about that distance? Doesn't light fall off with distance (inverse square law). I mean, this sort of confuses me. Sure, I'm pointing the camera at his face, but the light has to bounce off of his face and travel to the camera and through the lens and hit the film. With the reflected light meter, the light was already traveling that distance with my reading I was getting, with the incident it's not. How does this work exactly? Are there any good books/websites that explain this, it confuses me. You guys know how to explain this?
Another confusion comes from something I was trying to test out the other day. I set a 1K light in a room by the window, and went outside (it was night) with my camera. When taking the incident reading, would I hold the light meter right outside the window with the flat disc pointed at the window? Would I take the reading inside the house and put the disc right in front of the window the light is shining out of? But again, if I'm 100 feet from the house, the light still has to travel all that distance to get to the film and expose it.
Posted 15 October 2006 - 09:01 PM
Posted 15 October 2006 - 09:21 PM
When light strikes an object, it bounces in all direction. the only direction you need to worry about is the line that goes to the lens. Imagine it this way. You have 100 photons per nanosecond hitting a point in space (more would be needed for actual light, but this is a simplistic scientific view of what is going on). The surface is flat, so light would spread in one semi-sphere around that. Now its true that if you count the number of photons hitting a 1cmx1cm area, there would be greater incidence at one foot than at two. But your only concerend with how many photons there are in a single line (reflective) since that is the light that eventually gets focused onto the film plane. If in a single line there are 3 photons per nanosecond (all the others bounced in different directions) then there would be 3 photons per nanosecond at 1 foot and 3 photons at 20 feet.
A simpler way to put it...if you measured with your 8mm built in meter at 3 feet away, it would give you the same reading as if you were 10 feet away (assuming you have zoomed into the same area) Inverse law doesn't apply to the light that actually exposes the film. Only atmosphere (fog, very long runs through air, water) will affect the metering with an incident over distance.
Now apply that understanding to what an incident is actually telling you. All it says is that if x amount of light hits an 18% gray card, enough light will be reflected in any straight line to expose a film to medium density at n stop, assuming ASA and exposure time (frame rate) is constant. so if your trying to expose a face to key, then all you have to do is make sure your meter reads the stop your lens is set to when you point it at the light. Then you know that at just about any angle the camera could be at, enough light is transmitted from that point to the lens to expose the skin to key. I hope this helps. Your overthinking the incident meter, I am probably overthinking the response to it.
Posted 16 October 2006 - 01:36 PM
You are saying the inverse square law deals with the light sources and light bouncing off of them sources and not reflectance. Now, let me see if I understand it correctly in real life situations:
1) You are 10 feet away, filming directly INTO a light (the turned-on light bulb is your subject, for example). Would you hold the light meter right in front of the bulb, or hold it right in front of the camera? It seems to me that you would, in this circumstance, hold the meter in front of the camera since it's not reflectance, but direct light?
2) You have set up a light in a bedroom, and you shine it on a person standing in the bedroom. You go outside of the house and set the camera up 20 feet in front of the bedroom window. Now, lets forgot the time of day, as I expect during the daytime it would be hard to see because of light from the sun reflecting off of the window. But, in this circumstance, you would simply hold the light meter by the subject to see how much light is hitting him/her, get a reading, go outside and set the f/stop at that reading and everything will turn out ok? (or maybe open up just a little bit due to the CLEAN window being in the way?) Also, one would have to worry about the outside of the house being overexposed or underexposed, but we can forget about this right now.
3) You have set a light up in the bedroom, but the subject is again the light bulb. You go outside, 20 feet away from the window and set up the camera. When taking a light reading, do you stand right outside the window? right in front of the bulb? This one I'm a bit confused on....
I'm asking these lightbulb questions because I plan to film a scene in which everything is illuminated by house lamps. I was told that a reflecting light meter is better when dealing with items that emit light, such as a television.
This was my plan/my options. I'm not too sure on how to handle it yet, maybe you can help me out:
I want the whole scene to have an orangish tint to it. I will be shooting on Tungston film. I bought a few 200 and a few 300 watt incandescent bulbs. There will be two house lamps, so my initial plan was to use these powerful bulbs to illuminate the room, but as I thought about it more, there will be shots where the bulbs can be seen along with the actor. So now I deal with:
A) Using the 200 and/or 300 watt bulbs the entire time, and when the bulbs are in the shot they will be extremely overexposed, but when they are not in the shot it will have the nice look I'm going for (as my actor is walking through the room, having that color tint on his face)
Using 2 different types of bulbs, one more powerful and one weaker depending on what the shot will be. For tight shots where the bulb will be seen, along with the actor, use maybe a 100 watt light, or a 60 watt, and for the longer shots where the bulbs can't be seen, use the 300 watt. Now, the problem comes when the long shot has both the actor and the bulbs in it. It seems like I would then have to put the 60 watt bulbs in and use some tungston lights off camera with maybe a 1/4 CTO gel or 1/2 CTO gel to get the same look? This could cause some consistancy problems though if all tints didn't look the same.
pretty much my issues are now dealing with how to take readings of lights that are in the shot along with your subject.
Edited by Miles Sullivan, 16 October 2006 - 01:36 PM.
Posted 16 October 2006 - 02:23 PM
The reason is due to the difference in transmissive light and reflected light. Imagine how much light would have to fall on a white light bulb to equal the amount of light emitting from its surface. I did a test for a movie I am working on, and to get a peice of 216 to blow out (5 stops over) I only needed about 300fc of light on the back of the sheet. I was calibrated for 125 film at 24fps. If you do the math, after the light loss, an incident reading from the surface would give me around 2 stops over, but when I used a reflectance it was 5 stops over. The reason is incident assumes the light your metering must first strike an 18% gray object and be reflected back. Transmissive light doesn't reflect at all, it gets diffused by the bulb, but after that its projected straight through, so what your interested in is how much light is transmitting in one direction (reflectance)
1. Meter the bulb. If its too hot on camera, ND gell the camera side of the light (if possible) or increase the light with an off camera light to make the bulb fall into range. The reflectance reading of the bulb will not change regardless of camera angle or distance. Now this doenst matter if your taking a reading from 2ft away or 20, since its a reflectance reading. Just make sure if your using a reflectance from far away, that its a spot meter, otherwise if its a 45degree meter (like mine) then it will average the light of the bulb, pluss the reflection of the scene it sees. If you have a 45 degree, then you must be close, so the bulbs surface fills all 45 degrees.
2. Assuming its night (or you tent the window to prevent daylight reflection) you would use an incident and just set it that way (assuming the glass is 100% clear. adjust if its a window with ND on it) You can use your meter to determine light loss (clear glass is 99% transmisive or better, so I wouldn't compensate too much.)
If it is daytime outside, then the inside must match the brightness outside (opening up would only further blow out the daylight.) to adjust for this, you can tent the window, tent whatever is being reflected to reduce light on it(probaly difficult/impossible given the size of the reflection) or put a large scrim behind the camera, in between the window and what its reflecting. That way the reflected light is cut before it bounces off the window. Remember, its not the sun reflecting off the glass, but the sun reflecting off everything outside, that then reflects off the glass. At 400-500 fc direct sunlight and windows that reflect 80% or more from that side (more for energy efficient windows), that would greatly overpower most practicles you could have inside, and render the inside scene very hard to see (regardless of what you stop your lens to.)
Other options include building something dark into the window's reflection outside (position the reflection of a large tree in the reflected view where your actors face will be, etc) You can really trust your eyes here. Your eyes capture and translate light very similar to how film registers, the only difference is your eyes have a larger exposure latitude, so you will see far more detail in highlights and shadows than your film will. If there are a lot of reflections and variables, it would be best to place a gray card where your actors face will be and spot it from there (with whatever light is falling on him from inside hitting the card as well) That will show you how bright it will be with the reflection and face combined, but will not tell you nessisarily if the windows reflection will overpower your actors face. Your eyes can tell you that.
3. again if the subject is the lightbulb that you are concerned about, meter with a spot or reflectance and that will tell you where its going to fall in the camera, after accounting for glass light loss, reflection of the window, etc.
I have never been a big fan of practicles for the sole source of illumination. I usually shoot video, so sensitivity isn't the issue, its dynamic range, something film will help with of course. You may find however that using a bare or shaded light will blow out/flare the lens if its powerful enough to light your actors face. you may find its easier to get a bulb that will get to the right exposure range on camera with the stop you plan to shoot at, then augment that light with off camera light.
Also bulbs are hard to gell without them looking like they are gelled. A filter on camera or adjustment in the timing could help you avoid gelling an on-camera light. Also household bulbs are usualy slightly warmer than 3200k.