While I have always known exactly what my clients need from a photographic standpoint, I had to figure out how the entire input process worked in order to understand what my what my clients' needs were in digital terms. The mathematics below let me relate digital imaging to what I already know about film and photography so that I could determine exactly what I needed to do to make sure that I could offer my clients the right options. The Zone System teaches us that all films can record information within a range of reflectance and anything that's out of that range gets recorded as either maximum density or minimum density. In other words, there comes a point where a gray is so light or so dark that film can't tell the difference between that light gray and white or that dark gray and black. Input devices have the same limitation and it's called dynamic range. Dynamic range is usually written as a ratio in d log format. D log just means that the number they give represents a power of 10. So instead of writing that the dynamic range is 10 to the third power, they just write 3. By saying it's a ratio, they're just saying that the lightest value that the machine can separate out is 10 to the 3rd (10x10x10 or 1000) times brighter than the darkest value that it can separate. Now for the good part. We can mathematically compare a d log Dynamic Range with the Zone System. A Dynamic Range of 3 tells us that our brightest value is 1000 times brighter than our darkest value. Since every zone or stop represents a power of 2 (we're always doubling our brightness every time we open a stop), a dynamic range of 10 to the 3rd will roughly cover a 10 stop range of values (2 to the 10th power =2x2x2x2x2x2x2x2x2x2 = 1024 which comes pretty close to 10 to the 3rd or 1000). A device that has a dynamic range of 2.5 (10 to the 2 and 1/2 = 316), will cover only about an 8 stop range (2 to the 8th=256). The number of bits that the device uses to record an image determines how many values it can recognize between the two endpoints described by its dynamic range. The more bits the device uses, the more values it can see and the closer the image will look to continuous tone. Each bit corresponds to a power of 2, just like a stop. So if an input device uses 8 bits per channel, you'll get 2 to the 8th or 256 shades between the brightest value and the darkest value. If it uses 16 bits per channel, you'll get roughly 64,000 shades between those two endpoints. Now, the dynamic range and bits per channel of any device work together to determine what the final image will look like. If the device has a 1.5 dynamic range (5 stops), having 8 bits per channel (256 shades) divided over 5 stops might not be that bad--that's 51 shades per stop. With a dynamic range of 3 (10 stops), 8 bits gives only 25 shades per stop which would look too posterized for most commercial applications. Resolution simply refers to how many dots make up the picture and conceptually it's not that different from the idea of graininess in film. If there are too few dots that are too big, people will notice that the picture's made up of dots. With a large enough number of really small dots, no one will notice them. Unlike grain, each of the dots that make up a digital image has instructions associated with it that explain to the computer where the dot is located in the image, how bright it needs to be and what color it is. The more dots in an image, the larger the file size (the amount of memory it takes up) and the longer it takes a computer to do anything with it. Large files can be difficult to work with so the ideal input resolution will always be the minimum number of dots necessary for the final reproduction size of the image. There's a formula in image reproduction that states that the input resolution must be 1 1/2 to 2 times greater than the line screen at which the image will be printed when the image is reproduced at 100% or smaller. If the image is to be enlarged, the formula must be doubled every time the enlargement size doubles. In other words, if we scan a 4x5 transparency for 8x10 reproduction (a 200% enlargement) using a 150 line screen, we would need to multiply 150 by 2 for the formula and then by 2 again for the doubled enlargement to get the ideal input resolution of 600 dpi. Thus, the reproduction and line screen sizes that our clients use directly impact the input devices that will work for them. So, now that we have a way to compare devices, we can look at how our needs translate into digital terms. Our studio specializes in commercial still life and product shots all of which are done in the studio with total control over light and contrast. While we could control our lighting to reduce contrast and use a device with a smaller dynamic range, we are uncomfortable limiting our lighting styles just to conform with the limitations of an input device. We know our clients are happy with the way our film looks, and we certainly don't want to change that. We want to offer them the closest thing to film that digital technology can offer. Any input device we choose must have at least a 10 stop range (dynamic range of 3) and a minimum of 100 shades of grey per stop (10 bits per channel). We also know that our images are rarely reproduced much larger than 8 1/2 x 11 full bleed (9 x 11.5) with a 250 line screen. Therefore, we know that any input device we choose would have to be capable of image resolutions of at least 1125 dpi.
	Please note: the above article is © Judy Herrmann, 1995. It may be printed for easier reading but may not be distributed or sold without permission