Digital Imaging, Part 2: PPI, DPI & Resampling
Today's digital devices are more available than ever, and are very popular these days. Many are discovering how powerful and convenient the technology can be. However, digital technology isn't a solution if it isn't understood, and understanding the technology completely is far from easy.
In the last article of this series, PPI was introduced. Before the discussion continues, some clarifications and corrections concerning PPI must be made. PPI was incorrectly identified as the total number of pixels in a square inch. PPI, rather, is the number of pixels in an inch.
As such, every inch in a 12 PPI image is 12 pixels tall and 12 pixels wide. Therefore, a 12 PPI image will contain 144 pixels in every square inch (12x12 or 12^2=144).
Using this definition of PPI, a 300 PPI image contains 4-times the number of pixels in a 150 PPI image. This is because when you double the PPI, the pixel width and height of each square inch within the image is doubled.
For example, a 150 PPI image contains a total of 22,500 pixels (150x150 or 150^2). If you double the PPI to 300, you will find a 300 PPI image contains a total of 90,000 pixels (300x300 or 300^2). Note that if you multiply 22,500 by 4, you get 90,000.
Please remember that a digital image does not have physical dimensions by itself. PPI information gives an image physical dimensions by simply defining how many pixels equal an inch.
Similarly, a map may use a legend to indicate that every inch is equivalent to 1.2 miles. Without that legend, you have no way of knowing what the distance is (in absolute terms) between two points on the map. Without PPI, who's to say how wide or tall (in inches) a digital image is?
The terms PPI and DPI are often used interchangeably, which should be discouraged. PPI and DPI are similar in many ways, but are not the same thing. PPI always means "pixels per inch." DPI, in contrast, refers to "dots per inch." Many think of pixels as "dots" and equate the two terms. However, "dots" typically refer to dots printed on some sort of physical medium, like paper.
Furthermore, PPI should never be confused with a digital camera's "mega pixel" rating. Both measure the number of pixels in an image. However, PPI measures pixels in an inch, where mega pixel measures the total approximate number of pixels (in millions) in the entire image.
Like mega pixel ratings, PPI is often used as a measure of digital image quality. However, a very high PPI does not guarantee a good image. Nor does low PPI guarantee a bad image.
For example, suppose a photographer using a traditional film camera takes a picture using high quality film. The quality of the film does not guarantee a good picture. If the camera had scratched lenses or was simply out of focus, for example, the picture is likely of poor quality, regardless of what film was used. Similarly, a very high PPI is comparable to high quality film because although it has a greater potential to capture detail, it captures what the camera tells it to capture. Regardless, the higher the PPI, the more detail the image is capable of holding.
So what is considered low PPI and what is considered high PPI? Unfortunately, there isn't an easy answer - it depends on the situation. A detailed answer is beyond the scope of this article; however, 300 PPI is considered reasonable by many for average quality prints, especially if they're being printed on photo paper using a reasonably priced home ink jet printer.
Changing PPI alone does not alter the pixels in your image, or photo. Resampling, however, does alter pixels. Resampling simply means that you're adding or subtracting pixels from the image.
Any time you resample an image, you are also altering the resolution (the pixel dimensions) of the image. Removing pixels from an image is often called downsampling, and often used to bring overly large photos down to a reasonable size. When a photo is downsampled, the computer strategically selects pixels for removal. The photo is then recreated with the selected pixels removed.
When pixels are added, it is done using a process called interpolation. Interpolation "fills in" the gaps between the pixels that are created when the image is made larger. It does this by intelligently inspecting each pixel in the image and determining where pixels must be added, and what colors the added pixels should be.
In other words, when you make an image larger (in terms of pixel dimensions, not print sizes), it requires adding pixels to the image that aren't already there. As a result, the computer must intelligently "guess" what the image will look like at the larger size. Once it knows what the larger image should look like, it draws that image over the old one. This is also how the infamous digital zooms (in contrast to optical zooms) in modern digital cameras work.
Again, understanding all the facets of digital images (whether they're created from digital cameras, or scanners) is not an easy task. However, understanding the technology does require some knowledge of basic concepts. Without this knowledge, using digital imaging technology to its full extent is pretty tough.
In subsequent articles in this Digital Imaging series, digital images will be discussed in further detail and how the technology applies to scanners, printers and other digital imaging technologies.
Have comments or suggestions for a weekly Tech Tips article? Send an email to firstname.lastname@example.org.