A while back in 2000 after obtaining my first DSLR camera I realized that something was very different about the quality of the images I was getting. They were very grainy (old term for film) or noisy (todays digital term for grainy). I soon found that the way I was exposing the image that I saw in the viewfinder was much the same as the days when I was using slide films, meaning underexposure for color saturation. I even had much the same problems when I got the first DSLR camera in 2003.

Digital photography is very much different in the way of getting the correct or proper exposure. In a digital camera, unlike film, the camera’s digital sensor is made up of very tiny light gathering diodes called “photo sites”. Each of these sites have color filters generally in red, green and blue which are the primary colors. Each photo site receives light in varying amounts of brightness and color which in turns sends a varying electrical signal (analog) to an electrical converter in the camera’s processor. The processor then generates that signal into what we call a pixel. This is a very over simplified description but adequate to describe what I intend to convey.

The camera’s processor either sends this converted digital information as a JPEG, TIFF or RAW file format to a removable data card. You the photographer sets the file format in the camera’s menu depending on your choice of needs for the image. The camera’s processor produces this digital information in either 12 bit, 14 bit or 16 bits, depending on the camera type and model. This recorded information relates to the amount of tonal values that will be represented in the recorded image, based on your selection of image file format. Each of these file formats and the bit depth of the camera’s processor have a great deal of impact on the quality of the digital image file to work with. Modern DSLR cameras have multiple JPEG formats. However the JPEG file format is compressed and reduced in size and tonal quality of the image. JPEG’s are generally reserved for web/email use, and even the largest JPEG file is the least desired format as it is further reduced to only 8 bits due to in-camera processing.

Serious and Pro photographers obviously want the most quality for high end use in media reproductions or for large print sales and therefore set their cameras up to save the image file in RAW format. This format receives no image processing from the camera other than converting the electrical signal into a digital pixel. There is no image file compressing as that of the JPEG file. The RAW file format requires an extra step in post processing in an image editor such as PhotoShop...this step is called RAW Converter. The RAW converter gives the photographer a great deal of control over how the final image is to be rendered and to match what he or she saw at the time of composing and exposing the image in the field.

The remainder of this article is based on the RAW file format. Most of us by now know of the camera’s actual visual exposure scale known as the histogram, which is shown on the camera’s LCD panel on the back of the camera. The histogram is a visual scaled reference showing the actual placement of the darkest to lightest pixels, and more importantly, when those pixels are clipped or improperly exposed for that image. It was originally thought by most photographers that if the histogram’s visual scale was between the far left vertical line and the far right vertical line that the image was properly exposed. This would be quite true if it were for film but not necessarily true for digital.

Most modern pro-consumer and pro DSLR cameras are capable of capturing tonality in about 10 stops +/- of dynamic range, much more than that of yesteryear’s slide films BTW. It is the correct placement of where these tonal values are located within the histogram that will produce a properly exposed imaged. An image with no blocked highlights and shadows with the least amount of noise in the darkest shadows, depending on your composition. Ten stops of tonal values is generally sufficient for most landscape images, but there are situations where more tonal values are required. That leads into another separate article on bracketing 3 or more exposures to capture images with a higher dynamic range of tonal values generally referred to as HDR photography. There are many articles on HDR and how to best expose and process them.

As a visual reference of what 10 stops of tonal value would look like,see Figure One below. If you were to plot on a lined graph paper just how tonal values are distributed from left to right, you would visually see a vast difference in the percentage of pixels that would or should be to the right side of the middle of the histogram. The reason is that there is more tonal image information out there to capture than that of the left side of the histogram which represents the much darker tonal values. Lets say for this exercise that your 10 stop dynamic range digital camera captures say 10,000 tonal values, one in a normally exposed landscape image made in say mid morning. This is assuming the image was not a night shot or sunset. If you plot say 6 points evenly along a straight horizontal line from left (darkest) to right (brightest), you probably should see something like about half of those tonal values (approximately 5,000) on the first plot point just to the left side of the tall right vertical line of the histogram. Now take approximately half of those remaining tonal values or about 2500 for the second point on the graph to the left of the first point. Working your way left and evenly across the graph you would see the last point would be about 160 tonal values. This is graphically and not realistically correct but you see my point...there is no where near the amount of tonal values on the darkest side of the histogram as there would be on the brighter side. See Figure One, below.

This now leads us back to the beginning. If you have too much information showing up in the histogram on the far left side you are in great danger of capturing more noise than you intended. Yes, you can run noise reduction software in post image processing or even in your camera settings, but lets see if there is a better way of exposing so that noise becomes less of a problem. Besides, running in-camera noise reduction is processed at the same exposure time of the image capture and is performed at the end of exposure. This is exactly twice as long of a timeframe and cannot be performed if HDR bracket exposing.

We all have been preached that RAW captures more data (tonal value) than JPEG. Well, as mentioned above, that would be true in that the camera is not throwing out some range of tonal values during the compressing of the final JPEG image file. The RAW file is not capturing more...its just not processed like the JPEG. So therefore, yes you do have more tonal values in a RAW capture and the majority of those tonal values are out there in ‘Right Field’ so to speak.You also have that much more data (tonal information) to work with in post processing, especially in the RAW converter software.

So now lets think about pushing that intended landscape exposure to as far to the right of the histogram as possible without clipping it. Take that image file into the RAW converter and don’t panic just yet because you think you have lost all the detail in the shadows. Just pull the exposure slider back until you get the image back to where it looks good. Use all the controls in the RAW converter to tweak the image. However these are global image adjustments ( the entire image), much like those in “Lightroom”. If you are at all like me, global adjustments only get you half way to the final image. I generally don’t take my images too far using just global adjustments. I much rather prefer to work up my images in small increments of global adjustments. After the RAW converter adjustments and the image is brought into PhotoShop, I use multiple layers for local selections and feathering and further refinements that renders the final image. Some of my images will have up to 10-15 individual layers of image adjustments before I’m finished editing.

Now getting back to the correct exposure for a given landscape scene without harsh lighting of noonday sun. I first determine the limits of the composition left to right and top to bottom of what I want to show and then determine the limits of deep shadow detail required. Then I check the exposure for the brightest highlights and adjust the exposure to give me a histogram that just touches the far right vertical light. I might even cheat a bit and cross that line and see which areas in the image are blinking to determine the amount of highlight detail that might be lost for that exposure. You may have by now realized that I shoot nearly exclusively in manual mode. I would rather have to add some highlight detail that was lost than to deal with considerable noise in the dark shadow of the image.

Now as stated before, all of this information is assuming an evenly lit landscape scene and not specilityimages where the subject nearly fills the frame or in Fireworks or night photography.

Having read an article sometime back by Michael Reichmann of Luminous Landscapes about digital exposures, the OLE Light Bulb lit up quite bright. I went out and began experimenting with Michael’s technique and voila!...much improvement in my noise problem. Now this is a great tool but it does not entirely remove the threat of noise but it does reduce most of the headaches. Today’s modern camera processors also do a fantastic job of noise reduction, especially at high ISO, and we all find ourselves there from time to time.

This article here is a total watered down, and a very UN-technical version of Michael’s tutorial found on his website below. My attempt here is to bring this subject back to the forefront of those photographers that don’t know of this technique...mostly you old timers out there trying to expose your digital images like you did when dinosaurs roamed the earth and that strange cellulose material was still much the thing to use. Michael’s in-depth tutorial should and needs to be read, as should about a hundred others on this excellent site. I have followed Michael's site for years now.

Michael Reichmann, 2011 ETTR Tutorial