The creation and manipulation of images is a complex business, involving both technical and artistic considerations. Unfortunately, this document can only scratch the surface of such a vast subject.
All computer-generated images are represented on a monitor screen as a succession of picture elements or pixels, the intensity of which creates the illusion of a complete picture.
The software that creates on-screen images in the computer’s operating system (OS), although graphics applications often use another system, usually PostScript, which the OS must then interpret.
Graphics software comes in two groups: bitmap applications and vector applications.
A bitmap application, by definition, creates a bitmap image of a specified size, resolution and colour depth. This kind of program is ideal for working with material that contains continuous tones, as generated by a camera or scanner. Some programs, known as painting applications, can create such tones using artistic tools such as brushes and liquid effects.
This kind of software is sometimes called a drawing application, since it provides tools similar to those found in a technical drawing office. It employs mathematical descriptions, also known as object-orientated descriptions or vectors, to create each of the several objects within an image.
Apple’s iPhoto application, supplied with Mac OS X, lets you download images from a camera or scanner. You can then organise your pictures into collections or selected images to an audience in the form of a slideshow: the latter can be very effective if you have a large display or video projector.
Image Capture is a simple application for downloading images or sharing the output of cameras and scanners over a network. If you turn on your camera and select Enable Web Sharing in the Sharing tab, other Internet users can also look at the files in your camera: all they have to do is enter the Web address as shown in this pane.
Mac OS X also includes iDVD, which allows you to create multimedia presentations that can be viewed on any DVD player. These can include photographic slideshows created from your own collection of pictures.
If you’ve updated your QuickTime installation to QuickTime Pro you can also use QuickTime Player to create a slideshow. You should proceed as follows:-
Most major graphics applications, including Photoshop and Illustrator, can be enhanced by using plug-in files, which are placed a special folder with a name such as Plug-Ins.
This kind of application is sometimes known as a painting application, especially if the programme is designed for creating artistic material. It usually has toolbars and palettes, as shown below:-
This example has tools for straight or freehand lines, square-cornered or rounded rectangles, ovals and other polygons. The bezigon tool is particularly useful, since it creates Bezier curves using four control points or handles. The fill colour and pattern for all of these tools, together with pen width, colour and pattern can also be selected. And a brush and spray can are provided for other effects.
Finally, two special items worthy of mention are the eyedropper tool, which lets you choose a fill colour from the picture itself, and the magic wand, which allows you to select an area based on its colour.
100%expands each pixel in the image to the size of those on your screen, making a 360 dpi picture look four or five times bigger than its real size.
Every application has its own special shortcuts, although the following are common in the Mac OS:-
|Change ||Double-click |
|Magnify ||Double-click |
|Erase ||Double-click |
|Select ||Double-click |
|Shrink ||Hold |
|Create ||Press |
|Create ||Press |
|To ||Press |
To see how a picture would look when printed you must increase the screen magnification. If you’re using the Mac OS, which has a standard screen resolution of 72 dpi, with a 360 dpi printer, you’ll need a magnification of
360 dpi = 72 dpi × 5. Similarly, for an accurate view of an image to be printed at 300 dpi you should use a magnification of
416.66%, or the nearest possible setting.
Bitmaps are ideal for images that have detailed shading, grey tones or a high colour resolution. But if you adjust the scale of such an image it often gets distorted, since the pixels in the picture don’t align with those in your display or printer. Fortunately, most bitmaps can be scaled upwards by as much as
137% whilst retaining reasonable printing quality, although some people might argue over this figure.
You can usually use copy and paste on a bitmap image within a single application without problems. Similarly, you can use a clipping file inside one application in the Mac OS. However, when copying or pasting between different applications you may suffer from the following problems:-
Bitmap pictures that have been created using a digital camera are sometimes lacking in quality, often as a result of the compression employed in a JPEG image file (see below). The following procedure, described for Photoshop working under the Classic Mac OS, can render some improvement:-
Increasing the size of a bitmap results in distortion, mainly due to enlarged pixels becoming visible. Various mathematical methods can be used to fill-in the extra pixels to create an apparent increase in resolution. However, the results are usually blurred or suffer from some kind of unwanted artifacts. The following techniques are commonly encountered:-
Uses the colours of the nearest pixels to replace the missing pixels, usually resulting in jagged edges.
Employs a weighted average of the four closest pixels, blurring the image.
Use a bicubic function of the 4 by 4 closest pixels. This can be demanding on the computer and may suffer from ‘ghosting’ or produce jagged edges on diagonals.
A special technique, emulating natural patterns.
Uses the colours of 4 by 4, 6 by 6 or 8 by 8 pixels around the missing pixel.
All of these techniques can be improved by adaptive processing, in which the processing varies in accordance with changes across the picture.
This kind of software is also known as a drawing programme, since it creates images in the way you would in a technical drawing office, using tools such as a ‘flexi’ rule, set square and protractor. The toolbars in a vector application are usually a subset of those in a bitmap application, as shown below:-
Unlike a bitmap application, where each object is often merged into the existing image, you can move or modify each object in a vector application at any time. So, although a vector application may not be as artistically powerful as a bitmap program, it’s sometimes much easier to use and far more flexible.
You can select objects by clicking on them. The fill colour or pen attributes of the selection can be adjusted as required, or the selection can be moved, rotated or scaled. You can also reshape an item by dragging on it’s handles, and in some applications you can add extra handles to make this easier.
Objects can be grouped so that they behave as a single object, or ungrouped. They can also be locked to prevent important items being moved by accident. Objects can be moved to the front or placed behind other items, while advanced applications use layers to separate different elements of your work.
Vector images are defined mathematically, so they don’t distort when when you adjust their scale.
You can usually use copy and paste on a vector image within a single application without problems. Similarly, you can use a clipping file inside one application in the Mac OS. In addition, you shouldn’t have any problems copying images or using a clipping between two QuickDraw-based applications in the Classic environment or between a pair of Quartz-based applications in Mac OS X.
However, if you copy a image from a PostScript application into a non-PostScript application you’ll often get a bitmap picture. The reason is simple: non-PostScript applications don’t understand PostScript so they take what’s available, usually a PICT image of the correct size but at a resolution of 72 dpi.
Getting round this problem isn’t easy. One answer is to turn the graphic into a high-resolution bitmap. To do this, increase the screen magnification in your PostScript application: in the Mac OS you can get a resolution of 360 dpi with a setting of
360 dpi = 72 dpi × 5. Then use a screen capture utility to copy the enlarged picture, which you can paste into your non-PostScript application. Finally, you must scale the image to it’s original size, in this example by a factor of
Bitmap images that are added to a vector image document retain their original resolution, so they become distorted if resized. Worse still, documents containing such a bitmap are often unsuitable for high-resolution printing. You can also add a vector image to a bitmap image document, although it’s then a bitmap, often at 72 dpi and of an unpredictable size.
In theory, you can use vectorising to convert a bitmap into a vector image. GraphicConverter for the Mac OS can do this, although it takes a long time and can give unexpected results, usually because of complexities within the source material.
Numerous file formats are used, many of which are only recognised by their parent applications. However, the most common image files are understood by QuickTime, allowing them to be viewed from within any QuickTime-compatible application, including Preview (in Mac OS X) and Picture Viewer (in the Classic Mac OS).
TEXTand a creator code of
????. You can change these codes manually with a utility such as FileTyper: the files then gain the correct icons and should then open automatically in Picture Viewer.
As mentioned previously, a bitmap file can be very large compared to an equivalent vector image. This is especially true for pictures that have a greater colour depth, image resolution and size. The effect of size is particularly significant, as shown below for AppleWorks bitmaps of 360 dpi:-
|Size ||Size ||File |
|450 × 696||1.25 × 1.9||42 K|
|2000 × 2000||5.5 × 5.5||495 K|
Many applications use file compression to save disk space. Unfortunately, some programs don’t accept all types of compressed file, even when the file is a recognised type. Note that lossless compression doesn’t effect picture quality but lossy compression invariably removes a small amount of detail.
Some applications automatically add preview images or custom icons to saved documents. The preview images are shown in the Open dialogues in most applications, while custom icons appear on the desktop. The Open dialogues in the Mac OS often have a Show Preview button, letting you see the previews, and a Create Preview button, which adds a preview to a file that doesn’t have one.
The following two points refer to the Classic Mac OS:-
PICTresource in the document’s resource fork.
Some of the more common file formats are listed below, complete with filename extensions, shown in order of preference, and Classic Mac OS type codes.
This is the standard format for bitmap images in Windows and OS/2 systems on a PC, offering 1, 4, 8, 15 or 24 bits per pixel. A header inside the file specifies the image depth as follows:-
This kind of file is really designed for desktop publishing (DTP) applications. It’s similar to a PostScript file (see below), although the instructions aren’t in the usual PostScript language. Such documents can be used to store pages of vector images, bitmap images and text. Note however, that bitmap pictures are usually made up of TIFF data (see below) that’s embedded into the EPS file.
An ‘open standard’ for graphical images whose files can be opened quickly.
This bitmap format was devised by Compuserve for the Word Wide Web, accommodating 1 to 8 bits per pixel, giving a resolution of 256 colours or shades of grey. It uses lossless compression, which is particularly effective for synthetic images containing blocks of colour, line drawing or text.
GIF files come in two kinds: the original GIF 87a and newer GIF 89a. The latter accommodates comments, animation and transparency, but with only 255 shades. In theory, older applications might refuse GIF 89a files, although this is unlikely since this format appeared way back in 1989.
There are only 216 colours that can be easily displayed on any computer platform. These are known as Web-safe colours, all other shades requiring the use of dither, where pixels of different colours are alternated. Although dither is effective, it isn’t very attractive, as each of the pixels are clearly visible.
The Mac OS normally uses gamma correction, so as to match computer-generated colours more closely to the human eye. Unfortunately, different gamma settings are used in Windows and other computer platforms, which means that images created on a Mac can appear dark when viewed on other systems.
FKEYswitching device called GammaToggle (Richard Gustafsson).
The use of transparency within an image is useful where it’s placed over a coloured background. In this instance you should avoid anti-aliasing, as this causes unpleasant edge effects on the graphic.
The GIF 89a file format can contain multiple images for animation, complete with timing and loop information. Flip animations use a series of separate images whilst Cel animations use different layers, the lowest creating the fixed background with the remainder containing moving elements.
This proprietary format is used for photographic clip-art images, which have the backgrounds removed, allowing them to be easily inserted into other documents. The RGB colours are represented using 24 bits, with 8 bits in the alpha-channel accommodating 256 levels of transparency.
This kind of file is used by Illustrator for vector images. Apparently, the documents produced by Illustrator 9.x or higher are similar to a PDF, while older files are a simplified form of EPSF.
This format was devised by the Joint Photographic Experts Group (JPEG), a body set up by the International Standards Organisation (ISO) and the International Telegraph and Telephone Consultative Committee (CCITT). It employs sophisticated compression, allowing high-quality images to be kept in small files, which is ideal for sending such pictures over the Web. It supports 256 shades of grey or colour using 24-bits per pixel.
In some instances, lossless compression is used, reducing the file size by 5:1 or 6:1, although lossy compression is more common, giving a reduction of 10:1 to 30:1. The latter removes information that the eye can’t see and is particularly efficient for a colour resolution of 24 bits or more.
JPEG data is created by converting a RGB image into YUV form, in which
Y represents the luminance or brightness while
V define the colour. The latter two values have only half the resolution of
Y, exploiting the eye’s inability to resolve colour detail. The YUV planes are then divided into 8 by 8 pixel blocks that are Fourier transformed into fewer values of lower frequency. Variable Length Encoding is then used to minimise the data necessary for repetitive image patterns.
An Exchangeable Image File (EXIF) incorporates camera information inside its header. Most digital cameras use this to record the time and date the picture was taken, the exposure settings used, such as shutter speed and aperture, as well as other details, some of which may be unique to a particular type of camera.
Some JPEG files also contain IPTC data, as devised by the IPTC (International Press Telecommunications Council) and the Newspaper Association of America (NAA), consisting of text areas that identify the picture. These can be seen using the Info feature in graphics applications such as Photoshop and GraphicConverter.
This is the standard QuickDraw format, which can contain a bitmap or vector image. It’s used by most Classic Mac OS applications, as well as for storing images in the clipboard.
PICTand is available to the OS or other software. This kind of resource is also used for storing an image in a picture clipping file.
The PICT 1 format, used in the earliest Macs, accommodates monochrome images, although these can be dithered to improve their appearance. The PICT 2 standard, used in Macintosh System 6.0.5 or higher, employs 24-bit or 32-bit data, giving millions of colours or shades of grey.
PICTs use approximately 500 KB of disk space for a monochrome A4 page at 300 dpi, increasing to between 2 and 5 MB for shades of grey, and more for colour. Traditionally, PICTs use Run Length Encoding (RLE) for compression, although with QuickTime installed this is usually replaced by JPEG compression (see below). Fortunately, the JPEG-compressed data is stored inside a normal PICT header, so it can be used by any application, even those that aren’t aware of JPEG compression.
This bitmap file is used in the Classic Mac OS to provide a special picture on the screen during startup, replacing the Welcome to Mac OS graphics, as well as the second window with its progress bar.
You can use ResEdit (Apple), GraphicConverter (Thorsten Lemke) or other applications to create the file. It must contain a single
PICT resource that has an
0 and its filename must be StartUpScreen.
You can produce a startup screen file from any photo-quality image. However, a large picture is sometimes best avoided, since it has to be fitted exactly to the edges of the screen. Similarly, the appearance of startup icons on top of the picture may not be considered attractive. Finally, a larger picture occupies more disk space and takes longer to load. You can make your file as follows:-
Ideally, an image of
832 pixels wide should be
624 pixels high, giving an aspect ratio of
4:3, which matches the proportions of a traditional computer screen. Fortunately, you can fix any errors in height after initially setting the width. Supposing, for example, that your picture is
634 pixels tall, requiring you to remove
10 pixels, usually as
5 pixels from the top and bottom. Then you should proceed as follows:-
You can also use this to fix black edges, in which case you may have to use Scale Picture again.
This file format is old enough to form part of computer archaeology! Each document contains a 72 dpi black and white dithered bitmap image of 576 by 720 pixels, corresponding to an picture of approximately 8 inches by 10. Each file is automatically compressed to around 5 KB.
This kind of file contains an image with multiple resolutions, allowing you to easily zoom in an out. The data itself is compressed, but selective decompression lets you rapidly navigate around the picture.
A special format, as used in Nikon cameras.
This bitmap format, offering 1, 8 or 24 bits per pixel, is used in Photoshop and other applications. Images can be in greyscale, RGB, CMY, CMYK, HSL or HSV form, although some applications don’t support the last two variations. Photoshop can also save files in other generic formats.
This format, derived from PostScript, accommodates full-resolution desktop publishing (DTP) documents and is often used for images in Mac OS X. Apple’s Preview application, which is supplied with Mac OS X, lets you read and search through PDFs. However, some users prefer Adobe Reader (Adobe), formerly known as Acrobat Reader.
PDFs can be created in the Mac OS using Acrobat Distiller (Adobe) and other applications. In the Classic Mac OS you can also use PrintToPDF (James W Walker) or you can convert PostScript files into PDFs using GhostScript (Aladdin). Finally, there’s PDF-Blit (Kas Thomas), which works with the BBEdit text editor to create a PDF from plain text.
Recent versions of PDF files can incorporate layers. These can be used to provide text in different languages, the user switching between the text layers by clicking on labelled buttons, such as French and English. Layers can also be used for maps whose detail increases as you zoom in and reduces as you move out.
This modern bitmap format was designed to replace both GIF and JFIFs on the Web, although most Web sites continue to use the earlier GIF and JPEG formats. PNG files can accommodate images of 1, 2, 4, 8, 24 bits per pixel, although most come in PNG-8 (8-bit) or PNG-24 (24-bit) form.
This format also supports alpha-channel transparency, ensuring that background images appear correctly, and accommodates gamma correction, providing consistent colours on all computer platforms. Although PNGs can’t be used for animations, they’re can be 25% smaller than GIFs.
This is a plain text file for storing material in desktop publishing (DTP). The content, usually of vector images and text, is written in the PostScript language. However, most modern DTP applications use EPS files or PDFs, as described above, in preference to ‘raw’ PostScript documents.
A special format, employing standard QuickTime compression.
A less-common bitmap format, as used by Imagestudio (Letraset) and Painter, conveying grey, RGB, CMY, CMYK, HSL or HSV images. Some applications don’t support the HSL or HSV versions.
This proprietary vector format was devised by Macromedia for sending images and animations over the Web. SWF files can be created in applications such as Flash (Macromedia) or LiveMotion (Adobe).
Contains raw data that doesn’t comply to any standard file type. A special application, such as GraphicConverter, may be required to extract the image from such a file.
A vector format, designed for the Web and supported by several companies. These files contain plain text assembled as XML commands, operating in a similar way to PostScript. SVGs support ICC colour profiles, sRGB colour coding, up to 16 million colours, font kerning and ligatures. In addition, text can be arranged along any kind of path, such as a line or curve.
Since the data is written in XML, these files are more versatile than Shockwave Flash documents. For example, you can store SVG files in an XML-aware database and then search for every document written in the past month that was printed in a rectangular text frame with a blue background.
As used on Silicon Graphics computers, providing 256 shades of grey or 24-bit colour.
This machine-independent format, devised by Aldus in the late eighties, is often used for scanned bitmap images. TIFFs can be used for monochrome, greyscale, palette-based colour or full-colour images. RGB or CMYK encoding can be used, with up to 32-bit resolution in each colour channel.
32,767are standardised whilst those from
−32,787are assigned to developers, enabling them to create files that are acceptable to any application. Sadly, this often isn’t the case, although FreeHand and other modern applications can read most files.
The following classes of TIFF may be encountered:-
Contains a black-and-white image, using dither, if required. As usual, a 300 dpi image on a 72 dpi display often appears unsatisfactory, although it should print perfectly, as long as you don’t scale it.
This kind of image can be easily viewed or scaled. When printed, you’ll get a dithered picture at maximum resolution. Hence an 8-bit greyscale image at 72 dpi or 150 dpi looks good on a 300 or 600 dpi printer.
Used for the electronic distribution of images created on a Group 3 fax machine or equivalent device. These documents often have a filename extension of
.xif and may contain multi-page material that can only be viewed with a special application. Each dithered grey image has a resolution of 203 dpi horizontally and 98 dpi vertically, increasing to 198 dpi vertically in Fine mode. If you can’t use this format you can try the alternative
.efx format, which can be opened using eFax Microviewer.
This kind of file contains a colour look-up table, also known as a palette, that describes the set of colours used in the picture. Such an image can be easily viewed, scaled or printed.
This format uses three independent RGB 8-bit colour channels, providing over 16 million colours.
A file conforming to the TIFF 5 standard can use one of the following compression algorithms:-
|LZW ||Similar |
|LZW ||More |
|CCITT-3 ||Used |
This format, also known as True Colour, is often used in high-end PC applications. The file size can be reduced using RLE compression, although this option isn’t supported by some applications. A file of this type can be created in Mode 0, 1, 2, 3, 9, 10 or 11, and can employ 1, 8, 15 or 24 bits per pixel.
This format is used for icon graphics in the Windows operating system, is 16, 32, 48 or 64 pixels square and employs 1, 4, 8 or 24 bits per pixel. The 16 pixel version, with 4-bit colours, is also used for favorite icons or favicons, as found at Web sites and displayed by modern Web browsers.
A Windows format that can be used for a bitmap or vector image.
This kind of file is produced by X11, the standard graphical engine used in Unix and Linux, and also in Mac OS X, where X11-based applications can be used alongside standard Quartz-based software. The width and height of the bitmap image is recorded at the start of the file: this is followed by the image itself, which is described as an array of characters.
Specialised applications often use their own kinds of files. The most common files used in computer-aided design (CAD) include the following:-
Other common three-dimensional (3D) formats include:-
Conversion between all of these and other graphics formats can sometimes be tricky.
MacWorld magazine (UK), IDG Communications, 2001-2004
©Ray White 2004.