Bit Depth

The number of binary bits that define the shade or colour of each pixel in an image. For example, a pixel in a Black-and-White image has a depth of 1 bit, since it can only be white or black. The number of colour values a given bit depth can produce is equal to 2 to the power of the bit depth. For instance, a JPEG image has a bit depth of 24 and thus has 2²⁴ possible colour values (approximately 16.8 million). A GIF image has a bit depth of 8 (or less) and thus has 2⁸ or 256 possible colour values

CLUT

Colour Look-Up Table. The indexed colour reference (mapping) palette of a 256-colour (or less) indexed palette image. Each entry of the CLUT contains the 24 bit value for a single colour comprising 8 bits for each of Red, Green and Blue. Pixels within indexed colour images reference individual positions within the CLUT. If the CLUT is re-ordered without remapping of the image pixel pointers to the new positions of individual colours, all the colours within the image are messed up. In Corel PhotoPaint, The CLUT is synonymous with the Color Table, and also with the on-screen palette when viewing this palette set to Image Colors mode.

CMYK

A colour mode made up of cyan (C), magenta (M), yellow (Y), and black (K). In the CMYK colour mode, colour values are expressed as percentages, so a value of 100 for any ink means that it is being applied at full saturation. Used in most full-colour commercial printing, CMYK is like CMY, but the addition of black (K) allows for true blacks and a wider tonal range. The CMYK colour mode is based on the CMYK colour model.

Colour mode

A system that defines the number and kind of colours that make up a Bitmap image. Black-and-White, Grayscale, RGB, CMYK, and Paletted are examples of some popular colour modes.

Colour model

A simple colour chart that defines the range of colours displayed in a colour mode. RGB (red, green, blue), CMY (cyan, magenta, yellow), CMYK (cyan, magenta, yellow, black), HSB (Hue, Saturation, Brightness) , HLS (Hue, Lightness, Saturation), and CIE L*a*b (Lab) are examples of some popular colour models.

Colour palette

A colour palette is simply a collection of solid colours. In Corel PHOTO-PAINT, you can use the on-screen Color Palette, the Select Color dialogue box, or the Color Roll-Up to choose colours for fills, outlines, paper, and more. You can use standard colour collections like the Uniform Colour Palette, fully customizable colour palettes that you create and arrange, or colour matching systems like the PANTONE MATCHING SYSTEM. A custom colour palette is a fully customizable colour palette composed of up to 256 solid colours. You can choose, edit, and arrange the colours in your custom palette, then save the collection as a file with a .CPL extension. Custom palettes are useful for setting aside and organizing the colours that you use most often in your work.

Colour Table

A feature in Corel PHOTO-PAINT that views and edits colours in a paletted (indexed-colour) image. See also CLUT.

Colour space

A virtual representation of a device or colour model's colour gamut in electronic colour management. The boundaries and contours of a device's colour space are mapped by colour management software.

Colour value

A set of numbers that define a colour within a colour model. For example, in the RGB colour model, colour values of 255 for red (R) and zero for both green (G) and blue (B) will result in the colour red.

Dithering

Dithering is a method of simulating full-colour images on a 256-colour computer display. Pixels from the 256-colour palette are combined into patterns that approximate other colours. The human eye merges these patterns into a single combined colour when viewed at a distance. Dithering can increase the size of GIF images as it increases the likelihood of increasing the horizontal variation in such images. The use of the 216 colour Netscape palette (the non-dithering, browser default palette) avoids dithering. If necessary, this palette can be translated into either an adaptive or optimized palette for further colour reduction.

GIF (.GIF)

Graphics Interchange Format (pronounced 'jiff', not 'giff'). Originally developed by CompuServe in 1987, GIF is a graphic file format designed to take up a minimum of disk space and to be easily read and exchanged between systems. This format is commonly used for publishing images of 256 colours or less to the Internet. There are several variants of this file format.

The original specification, called 87a, allows for the storage of multiple Bitmap images into a single file, supports up to 8 bits per pixel (up to 2^8 = 256 colours), 4-pass interlacing, and compression using a variant of the LZW (Lempel-Ziv Welch) algorithm.

A more recent GIF specification, known as 89a, extends the 87a specification through the addition of transparency (for only a single CLUT entry (colour)), comment fields, and animation.

The compression method used in GIFs works by scanning the horizontal rows of the image and storing codes for sequentially-identical pixels in a data dictionary. Thus images with larger horizontal areas containing the same colour compress much better than those with greater horizontal variation or (for instance) horizontal colour gradients. Unfortunately this algorithm is patented (owned) by Unisys and royalties for its use apply. PNG (Portable Network Graphics) images use the open-source 'deflate' variant of the LZW compression scheme and therefore avoid patent issues and costs.

JFIF (.JPG)

JPEG File Interchange Format. This is the variant of the JPEG compression method used by most modern Web Browsers. See also JPEG.

JPEG (.JPG)

Established by the Joint Photographic Experts Group and pronounced 'jay-peg', this format is an international standard for compressed true colour and grayscale photographic images; it offers relatively high compression with little loss of image quality.

JPEGs use a lossy compression algorithm which works by converting the spatial image information into a frequency map. The compression algorithm uses a Discrete Cosine Transform (DCT) function to separate the high and low frequency information within the image. However high frequency information is lost first, and therefore as greater compression ratios are used more and more high frequency information is lost. At high compression ratios JPEGs exhibit 'chroma downsampling' effects which appear like jagged edges around sharp colour boundaries.

When viewed on computers with 8-bit (256 colour) displays, JPEGS are force dithered into an 8-bit colour space and the decompression of these images is usually slowed considerably. Also, the compression algorithm always treats image data as 24 bit data, so reducing these images to an 8-bit palette prior to compression frequently results in lower compression ratios and worse image quality, however this does vary with the particular compression variant (there are many) used in any particular software program (for instance, Photoshop's JPEG compression engine is quite different from that used in Corel's PhotoPaint).

JPEG compression also tends to introduce noise into solid colour areas at higher compression ratios. This produces distorted or blurred images artifacts into solid colour areas of the image and are thus not very good for pseudo transparency blending of backgrounds with images with similar edge colours.

JPEG is really just a compression method, not a file format. Thus, many different varieties of JPEG compression exist and new ones are continually being developed. Many proprietary compression methods exist which provide better and high compression ratios and the ability to JPEG-compress 32-bit CMYK, 32-bit alpha channeled, and 48 bit colour images. Web browsers use a variant of the original specification called JFIF (JPEG File Interchange Format). Another recent JPEG variant allows for the progressive display of JPEG images (p-JPEG, similar to GIF and PNG interlacing). This format is now supported by most modern web browsers and can actually reduce the overall file size of JPEG-compressed images slightly.

See also p-JPEG and JFIF.

Lossless

The maintenance of image quality of an image that has been compressed and decompressed. The process of compressing and decompressing often degrades image quality. A lossless image is one in which the image quality of a decompressed file appears nearly identical to the original.

Lossy

A noticeable degradation to image quality as a result of file compression. Choosing a high quality compression often results in very little loss of perceptible information. The lower the quality of compression, the poorer the image quality will be when the image is decompressed.

Netscape Navigator palette

An 8-bit palette of 256 colours used by the Web browser, Netscape Navigator. By using colours found in this colour palette, you can ensure that your image colours will display clearly on systems using this browser.

PNG

Portable Network Graphic image format (pronounced 'ping') is designed as a replacement for GIFs with several major advantages. These advantages include: the use of an open-source compression algorithm (the deflate variant of LZW), prefiltering of vertical patterns providing greater compression opportunities, improved 7-pass interlacing, support for both 24-bit true colour and 256 colour images, 64K levels of alpha channel transparency, and monitor gamma information storage ability amongst others. Full discussion on PNG images is reserved for another tutorial.

p-JPEG

Progressive JPEG image format. A recent variant of the JPEG compression algorithm use for true colour images. See also JPEG.

RGB

A colour mode that contains three components: red (R), green(G), and blue(B). The RGB colour mode is based on the RGB colour model. In the RGB colour mode, a value between 0 and 255 is assigned to each channel of red, green, and blue. An RGB colour with the component values 0:25:118, for example, contains no red, some green, and more blue, resulting in a slightly greenish blue colour. Monitors, scanners, and the human eye use RGB to produce or detect colour.

All images contain colour (even B/W ones !). The colour of an image can be defined using a colour mode. A colour mode is a system that defines the number and kind of colours that make up a Bitmap image. Black-and-White, Grayscale, RGB, CMYK, HSB, Lab and Paletted are examples of some popular colour modes. Each colour mode has an associated colour model, a simple colour chart that defines the range of colours displayed in a colour mode. RGB (red, green, blue), CMY (cyan, magenta, yellow), CMYK (cyan, magenta, yellow, black), HSB (Hue, Saturation, Brightness) , HLS (Hue, Lightness, Saturation), and CIE L*a*b (Lab) are examples of some popular colour models.

NOTE:

Please note that many of the above descriptions and definitions have been adapted from the Help files in Corel PhotoPaint and is (c) 1992 - 1998 Corel Corporation.