Choosing an optimum exposure for each photo you take allows you to record the greatest amount of digital information and gives the most flexibility when you develop the file.
Digital sensors don’t react to light the way film does and the technique for each is the opposite of the other. If you are programmed as a film photographer then properly exposed digital files can be counterintuitive.
When shooting film, especially transparencies where exposure is critical, I usually underexpose by a half to one full Exposure Value (1) below the light meters recommended setting. (2) That way I usually get a nicely saturated, colorful transparency.
Just the opposite holds for shooting in raw format (3) and exposing a digital sensor to light; if anything, you want to over expose creating a file that looks too light. The reason is that electric photo sensors record more and more digital data with more light. The more data you collect, the more flexibility and options you have for manipulations in post processing. Conversely, dark areas have much less data and do not develop as well. You clearly see this when trying rescue detail from dark shadows as you quickly go from a bit of recovery to a muddy looking dark gray tone.
Lighter areas have more data than dark zones not just because of the linear way that the sensor responds but exponentially because the difference between each Exposure Value is either twice or one half the amount of light. (4)
This technique is often called “exposing to the right” which refers to placing your exposure so that it favors the right-hand or lighter side of the histogram graph.
How on a practical level do you know how much to over expose a raw file and not end up making it so light that you clip important highlights? (5)
The most accurate method is to look at the camera’s histogram for the scene you are ready to shoot or have just taken. Many camera bodies can be programmed to display a histogram graph with the preview of the scene seen in Live View Mode. For photos that have been taken, the playback function can be set to display a thumbnail of the picture and its histogram.
The histogram is a graph with a horizontal and vertical scale. The horizontal scale displays black to white values with white on the right. It is made up of 256 tones from pure black to pure white.
The vertical scale shows the number of pixels in each of these tones. As a group they will form a curve whose shape will vary depending on the proportion of the various tonal values.
The camera histogram typically displays each of the Red, Green and Blue channels that make up a photo plus a composite white graph of the RGB together. Every time I take a shot, I look at the histogram displayed with a thumbnail of the photo on the LCD screen. If I’m set up on a tripod, I look at a Live View preview of the shot with its histogram.
If your camera does not have a histogram feature then bracket the shot favoring overexposure.
The histogram for an optimally exposed photo should have the R, G & B channel curves and the composite white curve well over to the right but not so far that their toes go beyond the right side of the graph. In fact, to preserve texture detail and color I keep the right end of the curve about one EV to the left of the graph’s right-hand side. If any of the curves go over the right edge, those areas will be blown out or clipped. Sometimes this is normal and desirable like scenes containing the sun or very bright spectral highlights off water.
The three pictures below are from the same scene.
The top one is from the LCD screen of a Nikon D810 camera showing a thumbnail of the scene and its separate RGB channels plus the white composite RGB.
The second photo is the original unprocessed raw file. It is too light but demonstrates how far you can go with a raw file and still recover data. My normal capture files are a bit darker than this.
The third photo is the final Lightroom processed file (a Virtual Copy that does not change any data in the original file and simply writes instructions to the main file.)
- 1) Exposure Value is a general term referring to either one aperture (f-stop) of light, one shutter speed or one ASA/ISO of film – sensor signal amplification. The difference between adjoining EV’s equals two times or one half the amount of light. The two times or one half depends on whether you are adding or subtracting light that hitting the film or sensor.
- 2) All light meters are designed to produce middle gray when you set your exposure by the middle, neutral or center reading. If you instead point the meter at pure white or complete black they become middle gray in the photo. Therefore, the best way to get a balanced exposure is to take the meter reading off an object or area in the scene that you perceive to be middle gray.
- 3) These exposure recommendations apply to shooting in raw file format not JPEG. Raw files contain far more digital data than JPEG plus JPEG has been processed by an agreed-upon set of development values. Raw is designed to contain lots of basic data and then be developed.
- 4) The math for comparing the amount of light when increasing Exposure Values over a range of 4 EV’s: 2 x 2 = 4 and 4 x 2 = 8 and 8 x 2 = 16 and 16 x 2 =32
- 5) Clipping refers to either over exposing or under exposing, so that either some of the lights go to pure white or some darks go to black containing no texture or color data.
.jpg)
.jpg)
.jpg)