You’ve taken a great shot and, naturally, want to produce as large a print as possible to display. But how can you know just how big you should make it?
Before we look at the main factors affecting this decision, it’s important to clarify the terms commonly used to describe image resolution so you have a clear concept of what they mean and where they are applied. Understanding these different measurement metrics will help you to understand when to use them, and give you some insight into how digital prints are produced.
Many people are unclear about the differences between pixels per inch (ppi), dots per inch (dpi) and lines per inch (lpi).
Pixels per inch (ppi) is used specifically to define digital image resolution. Each pixel represents a single hue (colour), saturation and lightness value for a single point in the image at a vertical and horizontal coordinate.
Dots per inch (dpi) represent the smallest physical size of the ink droplets produced by a printer on the output media (paper). Multiple dots are needed to reproduce a single pixel value.
Lines per inch (lpi) are used to define resolution for printers that are incapable of continuous tone printing. The term is sometimes substituted with line frequency, screen frequency or halftone resolution and covers rows of cells or arrays of smaller dots that make up a single tone.
Factors influencing print quality
1. The resolution of the image. Best results are always obtained when images are recorded with the highest image resolution and quality settings the camera provides. For any image, you can work out how large it should be printed using the pixel density in the print; in other words, the number of pixels/inch (ppi). (Inches are used to simplify calculations.) But first you need to know the printer’s output resolution.
2. The printer’s output resolution. Whereas digital images are made up of discrete blocks of colour known as ‘pixels’ the images produced by inkjet printers are produced by placing tiny dots of ink onto the paper. Photo printers typically use between six and 10 different hues and tones to reproduce colours and many ink dots are required to produce the hue and tone that corresponds with each pixel in the image.
All printer drivers contain RIP (raster image processor) software that describes the placement and size of the ink dots on the output media. The printer driver takes care of how the dots are actually laid down and usually limits the resolution it will handle to match the output capabilities of the printer’s printing head.
If the ppi resolution of the image sent to the printer is too high, the image will be interpolated down (downsampled) and image data is lost. If the image PPI is too low, it is interpolated up and there’s a real risk of interpolation artefacts (typically moirø©, banding and jagged edges) being introduced. For the best results, there should be a one-to-one correspondence between image ppi and the printer’s dpi setting.
3. Viewing distances and the limits of human vision. Human vision is good but it has its limitations. The average human eye cannot discern differences in resolution at output resolutions higher than about 350 pixels/inch with close examination.
However, the further you move back from the print, the less fine detail you are able to see, even in optimal lighting conditions. Consequently, it’s possible to reduce the output resolution without viewers seeing an obvious deterioration in image quality.
A typical photo inkjet printer provides a linear resolution of up to 1440 dpi, which means it can lay down 1440 droplets of ink in each linear inch of paper. These dots must be arranged into cells to reproduce the continuous tones in a digital photograph. Larger cells can contain more tones while smaller cells provide greater detail, so there’s a trade-off between colour handling and output resolution.
A popular rule of thumb says the optimal image resolution for printing at the most common output sizes should be 300 ppi. Using this rule, dividing the image resolution by 300 will tell you roughly how big you should make prints at snapshot and A4 sizes. (We’ll deal with larger output sizes later.)
Below 300 dpi resolution begins to fall off although, depending on the printer you use, you probably won’t see much difference in print quality between 250 dpi and 300 dpi without a magnifying loupe. At normal viewing distances these differences will probably be invisible so we’ll classify all resolutions above 250 dpi as ‘Excellent’.
If you use an 8-megapixel camera with a 3:2 aspect ratio, it should be possible to produce excellent borderless A4 prints at 300 dpi. The calculation works as follows: Divide the longer side of the image pixels (3504) by the length of the A4 sheet of paper in inches (11.69 inches). The result is 299.74 dpi, which is almost 300 dpi.
The same calculation used on higher-resolution images demonstrates that all cameras should be capable of producing excellent quality at output sizes of A4 and smaller. In fact, when printing at snapshot size, images from files larger than 1800 x 1200 pixels need to be downsampled.
Serious photographers like to make poster-sized prints of their best images. The same calculations can be used to estimate maximum output quality, taking account of the average distances from which different-sized prints are viewed.
Many photographers prefer to make larger prints with a white border around the image. The width of this border relative to the paper size is a matter of individual taste, although in most cases it’s at least an inch on each side.
The table below outlines the resolutions and output quality ratings for two popular enlargement sizes, allowing for a one-inch border along each shorter side of the image. It demonstrates clearly how different megapixel counts in digital cameras will support different degrees of enlargement.
At 180 dpi and lower output resolutions, there will be a noticeable loss of detail, particularly in highlights, and noise will be apparent in shadowed areas. This level would be classified as ‘Poor’ print quality.
Between 180 and 250 dpi there’s a grey area we have classified as ‘Good’. At the lower end of the scale, prints made with resolutions between 180 and 200 dpi may contain some artefacts in detailed areas but they probably won’t be visible in large prints at normal viewing distances.
Above 250 dpi, images will be good enough to withstand scrutiny at normal viewing distances. So what’s the ‘best’ viewing distance for enlarged prints?
Generally speaking, people naturally tend to take two standpoints, the first at a distance where they can comfortably take in the whole picture, and the second closer in where they can see details. The distances vary with individuals, but most people stand farther away from larger pictures and closer to smaller ones.
In practical terms, there’s a trade-off between print resolution and viewing distance. The table to the right shows the minimum resolution at different distances before the eye begins to see individual pixels in prints.
This explains why photos on billboards can be printed with very large dots, yet appear as continuous-tone images at the correct viewing distance.
Several other factors can influence how much an image should be enlarged. For landscapes, the eye usually expects to see detail to near its resolving limit, whereas images that show smooth surfaces and geometric structures can be less demanding.
This means you should be conservative when calculating how large to print landscape shots but will probably have more flexibility when printing photos in which the resolution of detail is less important. However, some regions in the less demanding subjects require better detail resolution than others. For example, hair and eyes in portraits usually need to be fully resolved, although skin can often benefit from some softening. These issues are usually addressed while editing shots, which are then printed for maximum resolution of the detailed areas.
The optimal viewing distance for images may also be influenced by textures within a photo and/ or the paper on which the image is printed. Detailed images usually look best (and can withstand closer inspection) on papers with smooth surfaces, whereas softer-looking shots can often benefit from being printed on textured papers and viewed from distances where the texture of the paper will complement the image.
Detailed images are also more likely to be subjected to close inspection because viewers usually want to scrutinise their intricacies. In contrast, it’s usually more satifying to view soft and diffuse pictures in their entirety to obtain an impression of the photographer’s intent.
Aspect ratio cropping
Borderless printing is only possible with some printers and usually involves cropping the image to fit the paper. Images with a 3:2 aspect ratio should fit entirely on a snapshot-sized print, whereas 4:3 aspect ratio images will be cropped.
Images taken with a 3:2 aspect ratio should fit entirely on a 6 x 4 inch (15 x 10cm) snapshot-sized print.
Cropping is inevitable when you print on most popular paper sizes and some printing software allows you to preview the cropping before committing to a print. Images with a 4:3 aspect ratio are cropped along the long sides, while those with 3:2 and 16:9 aspect ratios are cropped at the shorter sides. The red outlines in the illustrations below show the areas cropped with different aspect ratios.
Cropping a 4:3 aspect image.
Cropping a 3:2 aspect image.
Much more of the image is lost when a 16:9 aspect image is cropped for borderless printing.
When you want to preserve the entire image, you must allow for borders around the print. On A4 paper, the top and bottom borders will be wider than the borders along the short sides of the image.
This illustration shows one way of positioning a panoramic image on a sheet of A4 paper to avoid cropping and allow an attractive framing layout.
This is an excerpt from Photo Review Issue 56.