A complete graphic design, illustration, web design and typography course.

Articles in this section:

Our recommended reference books:

Scanning images

In recent years, scanners have changed from being a high technology device, highly priced, to turn up being one more peripheral of any domestic PC. These days, a basic colour scanner, capable to digitize up to an A4 size, may cost about 60 Euro / dollars, or less. Certainly, the scanner is a invaluable, even necessary tool for the designer. We can quickly acquire digitally photographies (or slides, if the machine is fitted for this option), drawings, prints or whatever image we want to use. Even one of the old hand-held scanners of those that were sold years ago, which were able to read a width of only about 10 cm. —only in black and white, is still an excellent tool. Many of the design works that we will do, don’t need anything else than a basic image in black and white, which later will further edited.

Resolution.

The resolution to which the scanner can acquire images is usually measured in dots per inch (DPI.) The typical values usually are between a minimum of 75dpi to a maximum of 1200dpi or more. There is a direct relation between the resolution of the image and its number of colours, and the size of the resulting image file after the scan. This size must be taken into account, to avoid unnecessarily big files, as they are more difficult to process and to send through the net. For the images that will be linked to a page Web, remember that the necessary resolution is only of 72 - 96 dpi; even with this little resolution (which matches the typical screen resolutions), we obtain images that, while having a small size, maintain all their quality.

Which resolution must we scan at?

En general, para imprimir al máximo de calidad las imágenes, suele recomendarse que se capten a una resolución doble de la del dispositivo de salida. Así podemos imprimir a tamaño real, e incluso aumentarlo un poco, sin merma en la definición. Es decir, para una impresora láser a 300dpi, podemos escanear a 600dpi. Pensemos que, inversamente, si tenemos una imagen a baja resolución, para imprimirla a mayor tamaño que el original se deberá escalar el mapa de bits y perderá calidad, con aparición de bordes dentados y mal aspecto general. En cambio, las imágenes para la pantalla no precisan resoluciones mayores que 72 dpi. Si inicialmente hemos trabajado con un mayor grado de detalle, al preparar la imagen definitiva para incorporar a la página web nos conviene bajar la resolución (con la opción de nuevo muestreo, resample o similar del programa). Más resolución no sirve para nada: la imagen no se verá mejor en el monitor (¡puede que no se vea de ningua manera, porque no todo el mundo tendrá la paciencia para esperar que baje! Como referencia práctica, esta tabla nos da una idea del tamaño en píxels que tendrá una imagen escaneada a 72 dpi. Podemos imprimirla y tenerla a mano al preparar escaneados para la web: Generally speaking, to print images to their maximum quality, usually it is recommended that they have a resolution doubling that of the output device. Thus we can print to real size, and even increase it a little, without definition loss. That is, for a laser printer at 300dpi, we can scan to 600dpi. Inversely, if we have an image in a low resolution, to print it to a size greater than the original, it will be necessary to scale the bitmap and it will lose quality, with an appearance of pixelated edges —a poor looking, non professional result.

However, images for the screen don’t need resolutions greater than 72 dpi, as we have commented already. If initially we have worked with a greater detail, when preparing the definitive image to incorporate to the web page we can decrease the resolution (with the resample option of the program, or similar). If you upload a higher resolution image to the web it will be wasted bandwitdh: the image will not look better on the screen (that is, if somebody ever sees it: probably they won’t have the patience to stand for an hour waiting for the image to finish downloading!) As a practical reference, the following table will give an idea of the size in pixels compared to its apparent physical size for images scanned at to 72. You can print it out and have it near at hand if you scan materials for the screen of the web:

Size (one dimension) of the original in centimetres (1cm = 2.54 in) Equivalent lenght in pixels, scanning at 72dpi.
1 cm 28 pixels
2 cm 57 pixels
3 cm 85 pixels
4 cm 113 pixels
5 cm 142 pixels
10 cm 283 pixels
15 cm 425 pixels
20 cm 566 pixels

We must pay attention to some facts. For example, an image that has a length of 20 cm, like the smaller dimension of a A4 sheet, would match the width of a VGA screen. In order to calculate the total size of the image —without compression— it depends on its colour model and its dimensions in pixels. The graphics programs usually indicate the expected size of the resulting image for a given format.

Using the scanner.

a scanned object

The operation of scanners is very simple. When you install the device, it usually bundles some kind of auxiliar program that allows to communicate the scanner and the graphic applications of your computer. Anyway, once the driver of your scanner has been installed, and the machine correctly set up, you may scan directly from most graphics software (from Photoshop, Paintshop pro, et cetera.) The program that controls the scanner is habitually called Twain Controller (or Twain-32.) When we access the graphics program, we have to choose an option in the menu like File|Acquire, Import|From scanner, or similar. This will let us digitize the image.

Many scanner drivers or interfaces offer first a quick preview of the image before actually acquiring it. In this preview we can select the part of the image that interests to us to scan, with a standard selection tool, and the type of scanning that it will peform (colour, grayscale, black and white, the resolution.) Once decided, the device reads the image and sends the digitized picture to the graphics program where we were working. Once within this application, we can carry out some adjustments to the image. The most common changes are to adjust its contrast, brightness, orientation and number of colors. The programs usually include specific commands that make it very easy.

The original (book, paper sheet) is placed on the open scanner and we just have to follow the instructions. When we say the original, it doesn’t necessarily mean paper. The scanner takes a king of photography: beams light and detects its reflection. Any object will be portrayed with good quality, although with reduced focal depth. The example of the toy “photo”, is really a toy that we have placed on the scanner glass. Some designers take advantage of this capacity of scanners to digitize textures and to photograph small objects. Some even place a transparent plastic film sheet (kitchen film can do) and cover it with cereals, grains of coffee, nails, fruit, pasta... to create one detailed macro photography-like texture.

Important points for saving digitized images.

At the time of saving the digitized image, we can choose between different formats. We must remember some compressed formats, like the JPEG, involve a certain quality loss in the image, due to the compression (the more compression, the more quality loss and undesired artifacts.) Thus, when opening the file again, we will find that it has worse quality than the original scanned image. For this reason, if we plan to come back to one image after publishing it, it is a good idea either to save it the less compressed that we can get away with, or save it in the native lossless format from the program (PSP for Paintshop Pro, PSD for Photoshop...) or a generic, cross-application file format that keeps the original quality of the image, like the tiff or png, with their lossless compression algorithm.

Tricks and ideas to process images.

To obtain better results when scanning in black and white, it is better not to choose the black and white option; it is preferable to use the grayscale option, which reads up to 256 tones of black, grey and white. After scanning and ajusting the picture, we can turn the image to 1-bit colour (black and white.) Why? Although an image is in “black and white”, actually, the ink or pencil on the paper displays darker and clearer areas, different tonalities. Thus, scanning in grayscale, we incorporated all this richness of tonalities; in black and white the scanner uses a tone threshold to discard points in grayish tones under this target threshold. In order to scan the subtle details of a pencil drawing, then, it is better to scan in grayscale and later, if it is necessary, to transform the image to black and white. Below you’ll find an example.

ejemplo: escanear en blanco y negro  

Notice how the quality of the image is better when it comes from an image scanned in grayscale than if it has been scanned directly with the black and white option.

If we want an image with solid areas of colour or black, we can fill the areas in the graphics program. This saves us some trouble filling up all the contours in the original paper drawing: it will be enough to define the basic areas and then fill them with the painting bucket or any other tool from your preferred painting or drawing program.

Drawing tablets.

A brief commentary to this type of devices, suitable for the digital art with programs in the style of Painter. The digitizer tablet is a peripheral which is sensible to the pressure and movement on its surface by a special mouse or a stylus, a plastic pen. This pressure-sensible “graphical pencil” will let you create effects similar to natural drawing and painting techniques. Some programs totally take advantage of these capacities, in special the already mentioned Painter (we recommended to try it; some of their demos (like the old 4 version) are save-disabled but never expire. The most important manufacturer of graphic tablets is Wacom; some basic models can be purchased from 80 - 100 euros - dollars. It is a very sensible investment for any designer who loves to draw and paint!

Therefore, this type of complement is perfect to draw and paint directly in the computer, and to take advantage of some underused features of painting programs. But tablets can also be very useful for technical drawing, in CAD and drawing programs where accuracy is very important.