Raster Image Size: Of Megapixels, Resolution, & Inches


Another 8-week session of my Photoshop Tools class at PNCA started last week. As I taught the first session it occurred to me that many of my past and present students could benefit from an article on image size and resolution.

The concepts of image size and resolution are often misunderstood. In order to even begin to discuss these concepts, however, we must first differentiate between the two basic methods of saving image data. A raster image is composed of pixels; the image is broken into a grid of dots and the color values of each dot are stored in the image. A vector image is composed of mathematical equations that can smoothly reproduce the image at any size. This discussion does not involve vector images, as they are infinitely scaleable and technically do not have “resolution.”

Pixel Dimensions

The only true measure of the size of a raster image is its pixel dimensions. This is because its size in inches is derived from its resolution. The pixels are “real;” they’re the actual data saved in the file. When you’re designing for on-screen use, it’s how you measure all your sizes.

Megapixels

The number of pixels in an image is often described in terms of megapixels. A megapixel is simply a million pixels; the difference between a 4MP image and an 8MP image is the 8MP image has twice as many pixels. A simple equation can be used to derive the megapixel measure of any image:

size in megapixels = pixel width * pixel height / 1,000,000

Image Resolution

Usually we deal with image resolution when designing for print. Resolution is a measure of pixel density; it literally means pixels per inch. To state it mathematically:

Resolution = Pixels / Inches

We use ppi, or pixels per inch, when referring to the resolution of the digital image. We use dpi, or dots per inch, to refer to dots of ink or toner placed on paper by a printer when printing. Generally, the higher the printer’s dpi, the better the print quality will be, as the device is capable of making smaller dots of CMYK inks and fitting them more tightly together. An image’s ppi will also directly affect print quality as well; we all too often see examples of low-resolution images in print.

Different printing processes require different optimum image resolutions. The typical requirement for high-quality printing is 300 ppi. In most printing workflows image resolutions higher than 300 ppi will not improve print quality, because at 300 ppi the printer’s ability to convert pixels to dots of ink has already been maximized. Always check your print job specifications to ensure that you meet the resolution requirements for that particular workflow.

Resolution is not as relevant when working with images for use on computer screens and hand-held devices; a computer screen simply displays the pixels in the image with no need to translate the information into dots of ink. While dimensions in print are usually specified in inches, we measure images by their pixel sizes when designing for use on the computer screen. The default resolution for onscreen images is 72 ppi.

The Conundrum of Image Size

Many designers have probably lost count of how many times their clients have asked them to take images from their website and use them for printed materials, and they’ve had to try to explain to their clients why this might not be possible. An image that looks like it’s 5″ wide on a computer screen might have 600 pixels in it (the exact size will depend on the resolution of that particular screen). However, it doesn’t take a math genius to see that:

while 600 pixels on screen might look about 5″ wide,
600 pixels / 300 ppi = 2 inches

So even though it looks fine on your computer screen at about 5″ wide, if you were to print the same image at 300 ppi, it would only be 2 inches wide. If you were to print it at 5″ wide, the resolution would drop to 120 ppi, and the quality would visibly suffer. If you used photoshop to resample it up to the required pixel width of 1500 pixels wide, you’d be adding a lot of interpolated pixels to the image, reducing its quality and blurring detail. The only really good solution is to find a larger version of that image that has enough pixels in it to get the job done.

Below are two slides I originally designed for my CS Fundamentals Workshop at PNCA that help demostrate these concepts. You can also view the entire presentation PDF.


,

No comments yet.

Leave a Reply