Let’s simplify!
I want to simplify because finding technical details is so-o-o easy, but finding understanding may be a tad more difficult for some.
First of all, Dots per Inch (DPI) and Pixels per Inch (PPI) are two ways of measuring resolution. DPI applies to printed images and PPI applies to images displayed on a monitor, screen, display, or whatever you want to call it. Other than that, the two terms are logically the same thing.
In this article I’ll refer to both of them as DPI (even though some people will complain that that’s wrong).
We need to consider this topic from several viewpoints, hence the sections below.
Images Stored in Files
Images can be stored in files on your computer’s hard disk. There are two types of images – raster and vector. Oops, I forgot, I’m supposed to simplify. So let’s call them dots (raster images) and curves (vector images). That better describes what is stored in the file. Dots are stored as, you probably guessed it, dots. Curves are stored as lines, curves, and fills.
This entire topic relates only to dot images (raster graphics). If you’re dealing with curves (vector graphics), you can stop reading now.
So you’re still with me, huh? I guess that means you’re dealing with a file full of dots, then. The file name might end with .tif, .gif, .png, .bmp, .jpg, or something else. It certainly doesn’t end with .svg, .eps, .cgm, .vml, .wmf, or .emf.
All of these files specify the number of horizontal dots and the number of vertical dots. Multiply those two numbers and you’ve got the total number of dots.
If you want to see these two numbers, right click on the file in Windows Explorer and left-click on properties. Then click on the details tab and look for dimensions. That will give you the number of dots horizontally followed by the number of dots vertically.
What about DPI? So far we’ve been talking about dots, not dots per inch.
Some of these file types also store the real-world size of the image in inches and/or the dots per inch. The DPI can be used to calculate the real-world size of the image (dots÷DPI) and/or the real-world size of the image can be used to calculate the DPI (dots÷inches).
So what? And I mean that with all my heart – SO WHAT? The real-world size of the image, in almost all cases, is irrelevant because we won’t be outputting it at its real-world size. We almost always resize it to something more appropriate to our purposes.
Real-world size and DPI are therefore nothing more than suggestions. Suggestions that we are free to ignore, which we usually do.
One final thought: Real-world size is particularly meaningless when we’re talking about photographs. It really means the size of the photograph, which has nothing to do with the size of the real-world objects that were photographed. How many heads have you seen that are ¼” high?
Outputting Images to a Printer:
The process of copying an image from a file to an output device is called rendering. (Yes, I know… but I’m attempting to keep this simple.)
You will use some kind of software to position and size the image on a virtual page, then you will send the virtual page to a printer driver, which may resize the image somewhat, then it goes to a printer, which may again resize the image somewhat. The aspect ratio and the actual number of dots may be changed throughout this process, too.
All of that sizing and resizing eventually renders the image as a number of dots horizontally and a number of dots vertically on a piece of paper. Divide those numbers by the corresponding real-world image size on the paper, et voilà, you have DPI. But it’s not the same DPI as is stored in the file; it’s the rendered DPI.
The rendered DPI might matter. The higher this number, the clearer the image might be. However it illustrates once again how unimportant the file’s DPI is.
So what? Once again with the So What, huh?
Are you really going to stop and calculate the rendered DPI? It’s more likely that you will look at the printed page and make a decision – is the image acceptable or not? Yes or No?
If the image is acceptable, you’re done. And you won’t give a second thought to the rendered DPI.
If the image is unacceptable, the problem can usually be traced back to the image in the file. However, notice that it is not the file’s DPI or the file’s real-world image size that is the culprit. Resolution issues like this come back to the number of dots, not the number of dots per inch.
Outputting to a Screen, Monitor, Display, Cell Phone, or Whatever:
This discussion is almost identical to the printer discussion above. However, there is one additional complication.
With a printer, you know (in most cases) how big the page will be. With a display, you have no idea. Everyone has a different size display. And to make it even more interesting, they can use Ctrl+ and Ctrl- to zoom that display. You have no idea what the real-world image size or the rendered DPI will be.
And if you can’t know, then it is totally irrelevant. You pays your money and you takes your chances.
Scanning an Image:
Now it matters!
When you scan an image, you are setting a real-world object (perhaps a photograph, a butt, or an important document) onto the scanner. It has real-world dimensions. And you are about to scan it into a file. The software needs to know how many dots to store in the file.
DPI is important because it tells the scanner how finely to scan the real-world object. If you specify 10 DPI, you’ll get a very small file size, but don’t expect a very good image. If you specify 9,600 DPI, you’ve probably gone overboard. Your file size will be massive. For general purposes, you may find 300 DPI acceptable. Increase it for more important scans. Increase it for images that you intend to render larger than the real-world object.
By the way, if you want to scan your butt, use a camera. You don’t want to break the scanner by sitting on it. (Then send me the picture.)
Conclusion:
In most cases, just ignore DPI and PPI. Scanning is the exception.
Test Yourself:
I scanned a photograph at 1,000 DPI and the size of the file it created was 1,294,538 bytes. The software I am using won’t accept images larger than 1,048,576 bytes, so I’ve decided to rescan the photo at a lower DPI. What DPI setting should I use?
Leave a Reply
You must be logged in to post a comment.