Bob Wade Starting with a Pixel
We start with the beginning, and in the beginning there was the pixel. The digital photographer (DP) said, "The pixel is good, from this little pixel we will make photographic images". To capture and to create the pixel the DP made the digital camera and the scanner. To display the pixel the DP created the monitor and the color printer. But the tiny pixel is still the basic building block from which all images are created and without an understanding of the pixel nothing else in digital imaging will make sense.
Pixel is just another way of saying "Picture Element". To the camera or scanner, it is just a very small portion of the total scene which the camera or scanner samples to determine its color and brightness and then stores this information away, to be used later to create an image. The more of these pixels we have and the smaller their size, the more detailed and accurate the image will be. This is just another way of saying that the resolution of the final image depends on the number of pixels captured.
The number of pixels is stated in different ways and this is sometimes confusing. For example, with a camera or scanner we can state resolution as a total number of pixels, as a number of pixels per inch, or as a dimensional measure such as "the image is 1200 pixels wide by 800 pixels high".
The total number of pixels in an image is not of much practical use. It is really useful only when comparing the resolution of different devices. For example, one camera captures 1.3 million pixels, another captures 2.4 million pixels. The second camera has a higher resolution but we must break this figure down into something useful before we can use the information to create images. We need to know the dimensional measurements in pixels to effectively use them to create an image. We will get back to this later but first let’s look at how the two most common hardware devices, the monitor and printer, use this pixel information to create pictures.
Monitors
The monitor uses electron guns to fire electrons at the screen which causes phosphors on the inside of the screen to glow a particular color. Monitors have limited resolution and the number of pixels required to produce an image is much smaller than required for a printer. If you want to know exactly how many pixels your monitor can display look at the display setting in control panel. The first figure displayed divided by the actual horizontal width of your monitor screen will give you the number of pixels per inch your monitor can display at that setting. If you don’t know how to do this or don’t want to bother right now lets just pick a figure between 70 and 75. It will be close and should illustrate the point.
My monitor is set to display 1024 x 768 pixels. If we use a "17 inch" monitor as an example we will see that it is actually a little over 12 inches wide. My horizontal resolution (1024) divided by the horizontal width of my screen (12 inches) equals about 85 pixels per inch. So, if my image were 850 pixels wide it should produce an image approximately 10 inches wide. If my image was 1700 pixels wide and I tried to display it on my screen it would be much wider than my viewing area. I would be able to see only a portion of it at one time and would have to scroll left or right to bring the remainder of the image into view. We have only discussed the horizontal dimension but the same procedure holds for vertical sizing.
Printers
Printers are a different story. Don’t confuse a printers "dots per inch" specification with pixels. Dots per inch measure the printer's resolution. It defines the size of each individual printed dot and how closely these dots can be spaced when printing. Telling your printer to print "300 pixels per inch" at a printer resolution of "1200 dots per inch" may seem confusing and contradictory but it isn’t. We are talking about two entirely different things. The pixel is huge when compared with the individual dots laid down by the printer and the printer will print multiple dots to form each pixel for two reasons.
First, since the pixel is usually larger than the printer’s dot, several dots may be required to recreate the pixel. Second, the pixels can be any one of millions of colors. The printer is capable of printing a very limited number of colors, usually 3 to 6. It creates the pixel’s color by printing a mixture of dots using its limited selection of colors to simulate the pixel color.
So what pixel resolution and printer resolution do we use? For printer resolution, the old saying "if some is good more is better" usually holds. If you want the best picture use a higher DPI resolution for your printer. For the image resolution, stated in pixels per inch, you will have to determine what is best for you. Most sources say that 300 pixels per inch is about optimum and little if anything is gained by going above that. Your personal standards may be different. If you are entirely satisfied with the pictures you get back from Wal-mart with your point-and-shoot camera you may want to try 150 pixels per inch. Specific numbers can also vary with the specific printers, ink, and paper used. Use the above figures as a starting point and experiment to find out just what your specific equipment can do and just what satisfies your personal standards.
Sizing of the printed output is also sometimes confusing. Basically it is the same as we discussed above for monitors. If we have an image composed of pixels with a pixel dimension of 1200 pixels by 900 pixels and our software instructs our printer to print this image at a resolution of 300 pixels per inch our printed picture will be 4 inches by 3 inches. If we choose 150 pixels per inch instead of 300, our picture would be 8 inches by 6 inches. It’s just simple division. Divide the total number of pixels you have by the number of pixels per inch (defined by the resolution you have selected) and the results are the dimension of the print. It only gets complicated when we refuse to recognize just how simple it is.
Pictures for email or for display on the "web".
Since this usually involves displaying an image on a monitor, the procedures we discussed for monitors usually apply here with a few additional considerations. First, we probably won’t be able to determine the optimum pixel resolution since this will vary with the capabilities of the monitor on the other end. An image may look exactly the right size on our monitor but be oversized for someone with is using a lower resolution on their monitor or it may appear smaller to someone with a higher resolution monitor. Try 70 to 75 pixels per inch to start and adjust as necessary.
Another thing to consider when sending images as attachments to email is the file size. No one enjoys twiddling their thumbs as a 7 or 8 megabyte file downloads when you could have accomplished the same thing with a 30 or 40 kilobyte file which would download in seconds. The exception to this would be when sending someone an image to be printed. It takes a lot of information to create a photographic image on a printer and this means a large file size. Just try to determine how the file will eventually be used and size it accordingly.
File Size
Digital images can consume huge chucks of memory and disk space. Just consider that an image with 30 bit color depth uses 30 bits to store the color information for each pixel. (What, we haven’t discussed color depth? That comes next.) If we have an image with pixel dimensions of 1200 x 900 (remember our 4 inch by 3 inch picture?) the file size can be determined roughly by multiplying 1200 x 900 to get the total number of pixels (which happens to be 1,080,000 in this instance). Multiply this by 30 to accommodate the information needed for color depth and divide the total by 8 (there are 8 bits in a byte) and we have a file that is a little over 4 megabytes. Not something you would want to send someone via email if they had a dial-up connection. It would also produce an image so large that it would be hard to view on some monitor screens. This is why image size should be appropriate for the intended use of the image. Having more data than you need for the purpose accomplishes nothing but wasting disk space.
Image sizing is accomplished using software (or firmware in your camera) during image capture or by resizing or resampling an image using your image processing software.
The file format used can also be used to control the size of an image. Some file formats compress an image by throwing away information it considers nonessential and some file formats limit the amount of color information stored which can reduce file size. For example, TIF files are lossless but are much bigger than a JPEG file. The smaller size of the JPEG is accomplished by discarding some information. There is no free lunch, each choice involves compromises. Evaluate your options and choose accordingly.
Color Depth
Color depth defines just how accurately a particular segment of your image can be recorded. Remember that the pixel is just a small sample of the total scene and that the only thing that is recorded is its color. The camera or scanner does this by recording the intensity of each of the three primary colors in it own "channel". If our color depth is 24 bits that means that each of the three channels is assigned 8 bits with which to define the color. The output from the three channels is combined to form the final color of the pixel.
Think of it this way, if you had a color depth of 3 bits you could have each of the three color channels either full on or full off. This means that you could have full red or no red, full green or no green, and full yellow or no yellow. These three colors can be mixed together but since the intensity or proportion of each color can not be varied, you can see that this provides for a very limit number of colors.
Compare this with a color depth of 24 bits. With 8 bits per color channel, each of which can have one of two values, we can define 256 different levels for each of the three colors. Since the value for each color can be combined with the color from the other channels, we can have over16 million colors