What is the best format for your prints?

What are raster and vector graphics, and how do they differ from one another? These are questions that stakeholders often posit when encountering image-related technics that don’t fall within their usual spheres of knowledge. The answers are, well, technical, but of interest to many in the world of textile printing. This blog post functions as a guide to help and illuminate those among our stakeholders who would like to learn more about image file types and about how using them correctly can lead to better products. We hope it will be of help.

The source image
The source image, i.e. the image of the customer’s logo that the printer needs to create the accompanying proof, is an important element for any textile printer. Not on account of its contours and color composition, but because of its file type. The question we have to ask ourselves and our customers every time a new image lands in our inbox is: Is it raster or vector?

Why is that? Because while raster files are usable as source image for textile prints, they are saddled with so many disadvantages that vector files are to be preferred in almost all cases. The fundamental reason why by far the most printers — Jet Sport among them — recommend their customers to submit images in vector format is the simle one that they yield better results. Why? Lets find out by examining the two formats’ advantages and disadvantages.

Raster
A raster image is modelled by a two-dimensional Cartesian coordinate system in which so-called ‘pixels’ are placed to together comprise the image. Each individual pixel is equipped with a color and a unique set of coordinates that precisely specifies its place on the grid. From there the pixel’s color engages in a kaleidoscopic collaboration with the surrounding colors, together comprising the complete image.

Any individual pixel is but an otiose, square color fragment. But place a multitude of these otherwise charmless quadrates into sufficiently fine-meshed patterns, and almost boundlessly beautiful, vivid, and elaborate mosaics can emerge. In the RGB color model – the color model raster files typically use – any given pixel can be equipped with one of no fewer than 16,777,216 different colors. Photorealistic images are always raster graphics, as co-acting pixels can produce compositions and color gradients no other image formats or color models can match. Another advantage with raster is that images are editable at a much more atomic level with software like Photoshop. It’s unusually tedious and time-consuming – and consequently almost never done – but with enough skill and patience a graphic designer can in principle edit a raster image all the way down at the level of the individual pixel.

The number of pixels of which a raster image is comprised determines its quality. This corollary bears the term ‘image resolution’ and is the deciding factor in how an image is interpreted by the retina. The more pixels, the clearer it appears (up to the eye’s biological limit, which – dependent on context – is about 300 PPI at a distance of about 2.5 feet). The density of the pixels is tallied as PPI or ‘pixels per inch.’ Let’s illuminate further with a concrete example:

If we want to print an image with a PPI of 240 – which is the lowest pixel density suitable for textile printing – onto a textile area with a width of six inches, we multiply the values and get: 240×6=1440. Thus, the raster image must have a width of 1440 pixels to be usable across the desired area. We can also move the variables around and determine the maximum size in which the raster image can be printed: If we have an image with a width of 1440 pixels and know the pixel density must not subceed 240, we divide the numbers and get: 1440/240=6. Hence, the raster image can be printed with a width of six inches before losing quality.

So far, so good. But the raster format’s weakness – and the knock-down argument against its employment for printing purposes – is its fixed pixel count. Once a raster image has been created comprising a certain amount of pixels, that fundamental composition is immutable. So if we find ourselves wanting to manufacture a logo with a width of e.g. eight inches – and want to keep its incumbent resolution – we require a larger pixel count than the image has.

Let’s again get out the calculator: 240×8=1920. We require 1920 pixels in the width for the new, larger image but have only 1440. So if we print the logo, it will emerge with a pixel count that fits an area 25% smaller than prescribed by the resolution. Each pixel is now stretched to cover a larger area and consequently appears duller and blurrier than before. Since logos often must come in varying sizes – the size of a logo on the back typically differs from that on the chest – a graphic designer needs to work with images that allow for easy handling and augmentation. This remains especially true if the logo contains a font that – as you can imagine – pays a disproportionate price for loss of quality. It doesn’t take many grainy pixels before a text ceases to be legible. For this reason all fonts on the internet – where users will often zoom in and out on a text they’re reading – uses the vector format.

A related factor that similarly works against the raster is the file size. An image’s pixel count is directly reflected in its size. Independent of PPI that – as you recall – is an expression of pixels per inch, the file size scales proportionately with the total pixel count. Two images with identical PPI but different dimensions will have different file sizes. Since the image editor software must be capable of handling data pertaining to an image’s every pixel, the file size will greatly influence the speed with which the program can process the image. The more pixels, the better the image quality. The better the image quality, the larger the file. The larger the file, the slower the processing and the quicker the hard drive fill-up. Size will additionally tend to be a retarding influence on the transfer of files from one system to another.

Vector
The vector file is based on mathematical formulas that calculate and draw precise geometric elements in a coordinate system. They are drawn through calculations of exact point values and their connection with lines and curves that together comprise the full graphical expression. Such geometric building blocks – lines, circles, polygons, curves etc. – make the vector file well-suited for use within structure-based line graphics, which typically deals with 2D images with clean features and homogenous, flat colors. The same building blocks simultaneously make the vector file unsuited for the realistic renderings often seen in digital photographs, as flat colors lack precisely the color depth crucial to image realism. An important characteristic of the vector file is that it – unlike the raster file – isn’t required to manage millions of pixels but only has to remember a few coordinate points and the line equations that connect them. Because of this the vector file takes up significantly less space and is easier to transfer, move around, and work with.

Let’s again illuminate with an example: If we want a circle to be rendered by a raster image, the image will have to manage several million binary digits (bits) and deal with the accompanying file size. A vector file, on the other hand, has it much easier. With just two digits, it first finds the circle’s center in the coordinate system and then determines its circumference by adding the radius value to the circle equation: (x-a)2+(y-b)2=r2. These values are all the vector file needs to draw the circle. Should we then want to enlarge it, the vector file simply replaces the radius value with a new one and draws the result of the new calculation. This much easier method requires very little disk space and permits infinite enlargement without concomitant deterioration of the resolution. The line edges will always look sharp and the image appear – and technically be – brand new.

That the image is redrawn every time it is altered or enlarged means that it can be divided into individual components. These can be separated, arranged, modified, and colorized as needed with just one or very few clicks. A graphic designer can make them any size, they can continually be assigned any Pantone color number, and their simpler compositions ultimately make them easier to apply to the textile. A vector file is accordingly much more versatile and easier to work with than a raster file. What it lacks in photo realism, it makes up for in responsiveness, and in an industry where looming deadlines are the rule rather than the exception, the importance of fast and flexible image editing cannot be overstated.

Identify your image’s file type
What file type is the image you hope to use as source for your logo? Find out by looking at the file extensions or by enlarging the image and seeing how it affects the quality. If it is worsened, it is very likely a raster file.

The most common raster file formats have the extensions .jpg, .gif, .bmp, .png, .tiff, .tif and .psd (Photoshop), while the most common vector file formats have the extensions .eps, .ai, .cdr, .svg, and .wmf/emf.

Summation
Let’s tie things up by recapitulating what we now know about raster and vector:

Both file types possess clear advantages. Raster is the preferred file format for realistic and colorful image renderings, while vector is the domain of the utilitarian, undemanding facsimile. Raster is used for photographs and is by far the most used image type on the internet, while vector is widely used in the design of logos with an eye toward marketing. Raster is luxurious, but demanding. Vector is ascetic, but smooth. Raster is a Gustave Courbet painting. Vector is an Albert Uderzo drawing. In the textile printing industry, option number two is preferred, reliably gifting us the shortest production cycles, the most inexpensive logos, and the happiest customers.

Thank you for reading.

‍