cgardner
Offline
Dedicated FM
Upload & Sell: Off
|
 p.2 #2 · Do I really need a grey card? | |
kylegehmlich wrote:
I just prefer grey because it's calibrated to work with my camera's meter's assumptions: that a reflective reading of 18% grey equals proper exposure. Using a towel can be fairly precise, I just trust calibrated tools more. (That is not a dig at anyone who uses the "towel method" or other similar method)
A reflective reading off 18% gray doesn't equal "proper" exposure. Most cameras are calibrated based on value of 12-13% falling in the middle of the scale, per ANSI standards. Also the speed on the menu of the camera (e.g. ISO 100) may not actually the true speed of the sensor, which is why ANSI calibrated hand held meters must be compensated to each camera body used to obtain optimum exposure.
Anecdotal accounts I've read say the 18% card originated in reproduction photography back in the 1920s as an exposure guide. Apparently even Kodak doesn't know exactly when it was first produced. Why 18%? Where a 50% dot would typically fall in a halftone. The white side of the card reflects 5x more light which where the highlights would typically fall. Back in the mid-1970s I worked making halftones and color separations by camera at National Geographic and the main process control tool, even then, was a gray scale used to evaluate dot placement so that account seems plausible to me.
There were many metering standards in the early days and it was assumed 18% was the mid-point of film sensitivity. Ansel Adams adopted the 18% card as the perceptual "Zone V" benchmark for his zone system which is part of why it became the Holy Grail of exposure. But in the process of developing ANSI standards it was discovered with testing that 12-13% reflectance was a more accurate benchmark for exposing negatives. Kodak considered changing the card, but Adams fearing that would somehow undermine the Zone System goldmine, reportedly camped out in Rochester and badgered Kodak until they relented and left it at 18%. Today if you buy a Kodak Gray Card Kit (Publication R-27) and read the instructions they say to adjust a meter reading off the card by 1/3 to obtain a more correct exposure. The cards are actually manufactured by Tiffen and are color neutral.
In any process control it best to measure what is most critical to the outcome. When shooting negatives exposure is based on putting adequate density in the shadow areas of the negative. For transparencies and digital the highlight detail is critical variable so the calibration of the exposure is best based on the highlights.
Compensation of a hand-held meter like a L-358 is performed done by taking a reading with the meter (e.g. f/5.6) then shooting a bracket series of exposures of test target like a white towel around that reading (e.g. from f/4 to f/8 in 1/3 stop increments) then looking at the resulting RAW files to evaluate visually, based on image detail, which exposure actually reproduces the textured highlights correctly. The towel makes an ideal target for that exercise because its easy to see in the loops of the fabric when overexposure starts to obliterate the detail.
A card can also be used to determine what ratios you're dealing with by placing it in various parts of the scene and taking readings off of it.
The ratio convention used for portraiture (e.g. 2:1, 3:1, 4:1) where the shadow value is constant describes the lighting on a face, not the entire scene and is based on several assumptions . The ratio indicates the relative REFLECTED brightness on the highlighted part of the face vs the shadow side (represented by the constant 1). The convention also assumes the key light overlaps fill which is the same (i.e. neutral) on both sides of the face. That is why there is no 1:1 ratio in the convention.
The earliest analog ratio meters consisted of a card with two holes, one clear placed over the shadow side of the face, and one a graduated ND filter. Each ND of .30 cuts light by a stop. By moving the ND over the highlight hole until it matched the brightness of the clear hole over the shadows the ratio of reflected brightness could be determined. Simple, clever, and very accurate.
When we meter two lights separately with an incident reading, the REFLECTIVE ratio isn't determined directly from the INCIDENT readings, it is inferred from the convention which assumes the key light overlaps fill that is even on both sides of the face.
If you take two identical lights and put them at identical distances, with fill over the camera and the key to the side in front of the face this is how the ratio math works:
H:S
1:1 one unit of fill hits the entire front of the face
1:0 equal strength key light overlaps the fill
===
2:1 because the light is additive 2x more light reflects from the highlights
In like manner a 3:1 is obtained when the key light is 2x (1-f/stop) brighter:
H:S
1:1 one unit of fill hits the entire front of the face
2:0 2x stronger key light overlaps the fill
===
3:1 because the light is additive 3x more light reflects from the highlights
In cinematography where many lights are used the convention is to express ratios between them in terms of incident strength. An incident ratio of 1:1 is actually the same as a reflected ratio of 2:1 in the portrait ratio convention. That is a source of confusion.
I guess a towel would work too, but I find that method a little cumbersome when you want several readings or your light changes often throughout a shoot. Plus you have to make actual exposures to use the towel, so you need to change your camera's settings every time you meter.
How you might use a towel to set ratios, if you dropped your meter and broke it, would be as follows:
If you want your fill at f/5.6 set your camera lens to f/5.6, put the towel where the face will be and adjust the light until the towel starts to clip. Turn off the fill light.
Change the camera aperture to the value desired for the key light (e.g. f/8 - one stop brighter than fill for a 3:1 reflected ratio) and adjust the light until the towel starts to clip.
The camera sensor becomes the meter, with the towel clipping the benchmark for exposure.
Ratios by the numbers are just a blueprint. What matters is how the light looks and whether or not it fits the context of the subject's age, gender, and the intended message. For example you might light a woman with various tone shadows depending on whether it was a portrait intended to be soft and feminine, or more serious and business-like (i.e. more stereotypically masculine).
The simplest way I've found to arrive and the desired ratio and perfect exposure with studio lighting is this:
1) Start with just the fill. Raise it until the image from the camera reveals the desired shadow tone overall. The lighter the foundation of filled shadows, the softer the final lighting will look.
2) Turn on the key light. Have the subject hold the towel next to their face (on the brighter key light side). Just raise the intensity until it clips, then back off a bit. You'll need to compare camera exposure warning with the RAW files initially to corollate the two.
The result is a full range of detail over the entire tonal scale that EXACTLY matches the range of the sensor, determined empirically by evaluating the results of the lights hitting the sensor.
The technical part, matching scene range to sensor for maximum use of its dynamic range, is really as simple as that. Take 5 minutes and try it and you'll see what I mean.... 
Making the technical stuff systematic and a no-brainer gets it out of the way and frees the mind for more important stuff like interacting with your subject...
Chuck
Edited on Oct 20, 2009 at 09:17 PM · View previous versions
|
