It’s a question many aspiring graphic designers find themselves asking: what is dpi and resolution? If you don’t yet fully understand dpi (dots per inch) image resolution: what it means, why it matters and how to work most efficiently with it, then read on…
First, we need to understand what type of images are affected by dpi level. In digital graphic design, we work with two types of image; ‘rasters’ and ‘vectors’. Simply put, a vector file is not an image; it is a list of mathematical instructions describing a collection of angles, curves and lines, which the computer reads and uses to draw an image on your screen. Each time you ‘zoom in’, move, or alter the image, the computer instantly redraws it, meaning it is not made from dots at all, and thus is unafected by dpi levels. No matter how close we ‘zoom in’, or how large we print out, we will not loose any visual quality. This is also known as being ‘resolution independant’.
The raster image (which is affected by dpi) is completely different; it is composed of a rectangular grid of evenly spaced coloured dots. The width and height of this grid form the physical size of the image, while the number of dots that fit within one inch of the physical size is the dpi (or ‘resolution’).
When we magnify or ‘zoom in’ on a raster image, at some point it will begin to look jagged and blocky – and these blocks are the dots which make up the image. If we have enough dots in our image, then at the normal viewing distance they will be so small that they can’t be seen – resulting in the clear image we want. If there are not enough dots, the software will enlarge them proportionally so they fill the physical size of the grid, resulting in a blockier, less clear image.
As a rule of thumb, designers traditionally aim to work at 300dpi – the standard setting used for many years in the print world. When designing for screen based work (websites etc) we usually work at the much lower 72dpi – the standard screen resolution of the Apple Mac computer (for reference, the PC operates at the slightly higher 96dpi). There will be times when working with rasters, when you will need to increase or decrease the size and/or resolution of an image; this is where things can, at first, be tricky to understand.
If we wish to make our image physically bigger, we can increase its size (make the image ‘rectangle’ bigger, but keep the number of dots the same). In Photoshop, go to Image, then Image size, and ensure the ‘resample image’ box is NOT ticked. In doing this, although the image itself is bigger (try printing it out to check), and each dot has grown slightly to cover the increased area, when storing the image on your computer, the size of its file will remain the same. Conversely, if we shrink the image, (again, without changing the dpi), the resulting smaller visual will retain its original file size.
You may find that when enlarging your image to the required size (thus lowering the dpi), the dots have had to cover such a large area that they become visible, causing the blocky ‘low reolution’ effect. It is here where we need to resize and resample the image. When resampling whilst enlarging (in Photoshop, go to Image, then Image size, and tick the ‘resample image’ box), we are effectively telling the computer to add new dots to the image, thereby increasing its dpi. In doing this, the software will analyse the dots present in the image, and add more – guessing what colour each new dot should be, based on the existing colours found immediately around it (this process is known as ‘interpolation’).
When increasing resolution & size, it is worth considering the resampling method used in the calculation. When in Photoshop, this can be selected at the bottom of the ‘image size’ dialog box. The drop down menu offers several methods of interpolation, each including information about the situation it is best suited for. Often, these differences will offer little if any visual difference, but selecting the appropriate method may help when atempting to enlarge a particularly low resolution image (This is not an issue when reducing resolution, where the software is removing information rather than adding to it).
Resampling an image to increase resolution can work wonders for an image, particularly when you already have a high resolution image, but need to increase its dpi to print it on a larger scale. It has its limitations however; try taking a small 72dpi web image via a Google search and enlarging/resampling it to A4 size at 300dpi. You may find that although the picture is now high resolution (when saving to disk, you’ll find its file size has grown considerably), the actual image is blurry and indestinct. The reason is that although the computer has added new dots to increase the resolution, the small number of dots in the original image meant it didn’t have enough information to guess what colour the new dots should be.
The more you work with images, the better you will become at predicting how well an image will survive enlargement. Its a combination not only of size and resolution, but colour, detail and subject – image resizing is as much art as science, and experience is the key, so keep practicing!