If you want to reharse around a trivial pursuit, use the PCtechguide glossary :

http://www.forums.pctechguide.com/glossary/bycat.php?catSelected=6&catSearchSubmit=View+Category

And then... looks for at least 6 mistakes and/or fuzziness

Category: Graphics (152 terms)

"Still" graphic terms explained (not multimedia)

TermDefinition
3D API A 3D application programming interface controls all aspects of the 3D rendering process. A mass of conflicting standards exist, including Microsoft’s DirectX and OpenGL, Intel’s 3DR, Reality Lab and Brender. Most are custom designed for either entertainment or serious 3D animation.
3D Graphics The display of objects and scenes with height, width, and depth information. The information is calculated in a co-ordinate system that represents three dimensions via x, y, and z axes.
Addressability Refers to how many pixels can be sent to the display horizontally and vertically. The most common combinations currently in use are 640x480 (VGA mode), 800x600 (SVGA mode), 1024x768, 1280x1024 and 1600x1200.
Aliasing A form of image distortion associated with signal sampling. A common form of aliasing is a stair-stepped appearance along diagonal and curved lines.
Alpha Additional colour component in some representations of pixels, along with red, green, and blue (RGB). The alpha channel denotes transparency or opacity, often as a fractional value, used in blending and anti-aliasing.
Alpha Blending An approach which uses the alpha channel to control how an object or bitmap interacts visually with its surroundings. It can be used to layer multiple textures onto a 3D object, or to simulate the translucency of glass or mask out areas of background.
Alpha Channel The extra layer of 8-bit greyscale carried by a 32-bit graphic. This extra information is used to determine the transparency or edge characteristics of the image.
Anamorphic Unequally scaled in vertical and horizontal dimensions.
Anti-aliasing Hides the jagged effect of image diagonals (sometimes called jaggies) by modulating the intensity on either side of the diagonal boundaries, creating localised blurring along these edges and reducing the appearance of stepping.
Artefact Unsightly visual side effect caused by defects in compression or other digital manipulation. Common artefacts include jaggies, polygon shearing (where 3D objects are torn or warped when screen refreshes can't keep up with 3D activity) and pixelation (where texture maps lose resolution and look blocky close up).
Asymmetrical Compression A system which requires more processing capability to compress an image than to decompress an image. It is typically used for the mass distribution of programs on media such as CD-ROM, where significant expense can be incurred for the production and compression of the program but the playback system must be low in cost.
Back Buffer A hidden drawing buffer used in double-buffering. Graphics are drawn into the back buffer so that the rendering process cannot be seen by the user. When the drawing is complete, the front and back buffers are swapped.
Bezier A way of mathematically describing a curve, used by graphics programs such as MacroMedia FreeHand and Adobe Illustrator.
Bi-linear Filtering Improves the look of blocky, low-resolution 3D textures when viewed close up by blending and interpolating groups of texels to create a smoother image.
Bit Depth In colour images, the number of colours used to represent the image. Typical values are 8-, 16-, and 24-bit colour, allowing 256, 65,536 and 16,777,216 colours to be represented. The latter is known as true colour, because 16.8 million different colours is about as many as the human eye can distinguish. Also referred to as colour depth.
Bitmap File File in which every pixel on screen is represented by a piece of data in memory, usually graphics although some audio formats are described as bitmapped as well. As opposed to a vector image, in which only a description of the image is stored. Each pixel can be represented by one bit (simple black and white) or up to 32 bits (high-definition colour). Uses the file extension "BMP".
Blockiness The consequence of portions of an image breaking into little squares due to over-compression or a video file overwhelming a computer’s processor. See also Artefact.
BLT Bit-aLigned BLock Transfer: the process of copying pixels or other data from one place in memory to another.
BPP Bits Per Pixel: the number of bits used to represent the colour value of each pixel in a digitised image.
Brightness A measure of the overall intensity of the image. The lower the brightness value, the darker the image; the higher the value, the lighter the image will be.
Bump Mapping A 3D rendering lighting technique designed to give a texture a three-dimensional, animated feel.
Camera In 3D graphics, the viewpoint through which a scene is viewed. Flythroughs of scenes are conceptually a moving camera.
CGA Colour Graphics Adapter: a low-resolution video display standard, invented for the first IBM PC. CGA's highest resolution mode is 2 colours at a resolution of 640 x 200 pixels.
Chroma The colour portion of a video signal that includes hue and saturation information. Requires luminance, or light intensity, to make it visible. Also referred to as Chrominance.
CIE Commission International de l'Eclairage: the international organisation that establishes methods for measuring colour. Their colour standards for colourmetric measurements are internationally accepted specifications that define colour values mathematically.
CIELAB (L*a*b*) A colour model to approximate human vision. The model consists of three variables: L* for luminosity, a* for one colour axis, and b* for the other colour axis. CIELAB is a good model of the Munsell colour system and human vision.
CIELUV (L*u*v) A colour space model produced in 1978 by the CIE at the same time as the L*a*b model. CIE L*u*v is used with colour monitors, whereas CIE L*a*b is used with colour print production.
Clip Art A collection of icons, buttons and other useful image files, along with sound and video files, that can be inserted into documents.
Clipping Removing, from the processing pipeline to spare unneeded work, complete objects and surfaces which are outside the field of view (known as the "viewing frustrum"). Also known as Culling.
CMYK Cyan, Magenta, Yellow, Black: the four process colours that are used in four-colour printed reproduction.
Colour Balance The process of matching the amplitudes of red, green and blue signals so the resulting mixture makes an accurate white colour.
Colour Palette Also called a colour lookup table (CLUT), index map, or colour map, it is a commonly-used method for saving file space when creating colour images. Instead of each pixel containing its own RGB values, which would require 24 bits, each pixel holds an 8-bit value, which is an index number into the colour palette. The colour palette contains a 256-colour subset of the 16 million unique displayable colours.
Compound Document A file that has more than one element (text, graphics, voice, video) mixed together.
Continuous Tone An image that has all the values (0 to 100%) of grey (black and white) or colour in it. A photograph is a continuous tone image.
Contrast The range between the lightest tones and the darkest tones in an image. The lower the number value, the more closely the shades will resemble each other. The higher the number, the more the shades will stand out from each other.
DAC Digital-to-Analogue Converter: a device (usually a single chip) that converts digital data into analogue signals. Video adapters require DACs to convert digital data to analogue signals that the monitor can process.
DCT Discrete Cosine Transform: used in the JPEG and MPEG image compression algorithms. Refers to the coding methodology used to reduce the number of bits for actual data compression. DCT transforms a block of pixel intensities into a block of frequency transform coefficients. The transform is then applied to new blocks until the entire image is transformed.
Decal A texture that is placed specifically on one part of a 3D object.
Decal Alpha Texture Blending Decal texture blending with the addition of alpha transparency.
Decal Texture Blending A blend technique where the triangle colour at each vertex is strictly the colour of the texture. The triangle's colour doesn't alter the texture colour.
Density The degree of darkness of an image. Also, percent of screen used in an image.
Depth Cueing Used in conjunction with fogging, depth cueing is the adjustment of the hue and colour of objects in relation to their distance from the viewpoint.
DIB File Format Device-Independent Bitmap Format: a common bitmap format for Windows applications.
Dithering The process of intentionally mixing colours of adjacent pixels. Dithering is usually needed for 8-bit colour, and sometimes for 16-bit. It allows a limited colour set to approximate a broader range, by mixing groups of varying-colour pixels in a semi-random pattern. Without dithering, colour gradients like sky or sunset tend to show "banding" artefacts.
DXF Drawing Exchange Format: the industry standard 3D data format.
EGA Enhanced Graphics Adapter: the IBM standard for colour displays prior to the VGA standard. It specified a resolution of 640x350 with up to 256 colours and a 9-pin (DB-9) connector.
Extrusion Taking a flat, 2-D object and adding a z plane to expand it into 3-D space.
Flat Shading The simplest form of 3D shading which fills polygons with one colour. Processor overheads are negligible and 3D games will allow the graphics to be stripped down to flat shading to improve the frame rate.
Fogging The alteration of the visibility or clarity of an object, depending on how far the object is from the camera. Usually implemented by adding a fixed colour (fog colour) to each pixel. Also known as Haze.
Fractals Along with raster and vector graphics, a way of defining graphics in a computer. Fractal graphics translate the natural curves of an object into mathematical formulas, from which the image can later be constructed.
Frame Buffer Display memory that temporarily stores (buffers) a full frame of picture data at one time.Frame buffers are composed of arrays of bit values that correspond to the display’s pixels. The number of bits per pixel in the frame buffer determines the complexity of images that can be displayed.
Gamma A mathematical curve representing both the contrast and brightness of an image. Moving the curve in one direction will make the image both darker and decrease the contrast. Moving the curve the other direction will make the image both lighter and increase the contrast.
Gamma Correction A form of tone mapping in which the shape of the tone map is a gamma.
Gamma Curve A mathematical function that describes the non-linear tonal response of many printers and monitors. A tone map that has the shape of this its compensating function cancels the nonlinearities in printers and monitors.
Gamut The range of colours that can be captured or represented by a device. When a colour is outside a device's gamut, the device represents that colour as some other colour.
GDI Graphics Device Interface: the component of Windows that permits applications to draw on screens, printers, and other output devices. The GDI provides hundreds of convenient functions for drawing lines, circles, and polygons, rendering fonts, querying devices for their output capabilities, and more.
Geometry The computation of the base properties for each point (vertex) of the triangles forming the objects in the 3D world. These properties include x-y-z co-ordinates, RGB values, alpha translucency, reflectivity and others. The geometry calculations involve transformation from 3D world co-ordinates into corresponding 2-D screen co-ordinates, clipping off any parts not visible on screen and lighting.
GIF Graphics Interchange Format: an image used by CompuServe and other on-line formats. Limited to 256 colours but supports transparency without an alpha channel and animation.
Gouraud Shading A method of hiding the boundaries between polygons by modulating the light intensity across each one in a polygon mesh.
Gradient In graphics, having an area smoothly blend from one colour to another, or from black to white, or vice versa.
Graphics Card An expansion card that interprets drawing instructions sent by the CPU, processes them via a dedicated graphics processor and writes the resulting frame data to the frame buffer. Also called video adapter (the term "graphics accelerator" is no longer in use).
Graphics Library A tool set for application programmers, interfaced with an application programmer’s interface, or API. The graphics library usually includes a defined set of primitives and function calls that enable the programmer to bypass many low-level programming tasks.
Graphics Processor The specialised processor at the heart of the graphics card. Modern chipsets can also integrate video processing, 3D polygon setup and texturing routines, and, in some cases, the RAMDAC.
Greyscale Shades of grey that represent light and dark portions of an image. Colour images can also be converted to greyscale where the colours are represented by various shades of grey.
GUI Graphical User Interface: a graphics-based user interface that incorporates icons, pull-down menus and a mouse. The GUI has become the standard way for users to interact with a computer. The first graphical user interface was designed by Xerox Corporation’s Palo Alto Research Centre in the 1970s, but it was not until the 1980s and the emergence of the Apple Macintosh that graphical user interfaces became popular. The three major GUIs in popular use today are Windows, Macintosh and Motif.
Halftone An image type that simulates greyscale by varying the sizes of the dots. Highly coloured areas consist of large colour dots, while lighter areas consist of smaller dots.
High Colour Graphics cards that can show 16-bit colour (up to 65,536 colours).
Highlight The brightest part of an image.
HSB Hue Saturation Brightness: with the HSB model, all colours can be defined by expressing their levels of hue (the pigment), saturation (the amount of pigment) and brightness (the amount of white included), in percentages.
Hue The attribute of a visual sensation according to which an area appears to be similar to one of the perceived colours, red, yellow, green and blue, or a combination of two of them. Also referred to as tint.
Image The computerised representation of a picture or graphic.
Image Resolution The fineness or coarseness of an image as it was digitised, measured in Dots Per Inch (DPI), typically from 200 to 400 DPI.
Interpolation The process of averaging pixel information when scaling an image. When the size of an image is reduced, pixels are averaged to create a single new pixel; when an image is scaled up in size, additional pixels are created by averaging pixels of the smaller image.
Jaggies Also known as Aliasing. A term for the jagged visual appearance of lines and shapes in raster pictures that results from producing graphics on a grid format. This effect can be reduced by increasing the sample rate in scan conversion.
JPEG Joint Photographic Experts Group: supported by the ISO, the JPEG committee proposes an international standard primarily directed at continuous-tone, still-image compression. Uses DCT (Discrete Cosine Transfer) algorithm to shrink the amount of data necessary to represent digital images anywhere from 2:1 to 30:1, depending on image type. JPEG compression works by filtering out an image's high-frequency information to reduce the volume of data and then compressing the resulting data with a compression algorithm. Low-frequency information does more to define the characteristics of an image, so losing some high frequency information doesn't necessarily affect the image quality.
Just-Noticeable Difference In the CIELAB colour model, a difference in hue, chroma, or intensity, or some combination of all three, that is apparent to a trained observer under ideal lighting conditions. A just-noticeable difference is a change of 1; a change of 5 is apparent to most people most of the time.
Lathing Creating a 3-D surface by rotating a 2-D spline around an axis.
LFB Linear Frame Buffer: a buffer organised in a linear fashion, so that a single address increment can be used to step from one pixel to the pixel below it in the next scan line in the frame buffer. The entire LFB can be addressed using a single 32-bit pointer.
Lighting A mathematical formula for approximating the physical effect of light from various sources striking objects. Typical lighting models use light sources, an object’s position & orientation and surface type.
Line Art A type of graphic consisting entirely of lines, without any shading.
Lossless A way of compressing data without losing any information; formats such as GIF are lossless.
Lossy A way of compressing by throwing data away; this results in much smaller file sizes than with lossless compression, but at the expense of some artefacts. Many experts believe that up to 95 percent of the data in a typical image may be discarded without a noticeable loss in apparent resolution.
Luminance The amount of light intensity; one of the three image characteristics coded in composite television (represented by the letter Y). May be measured in lux or foot-candles. Also referred to as Intensity.
LZW Lempel-Zif-Welch: a popular data compression technique developed in 1977 by J. Ziv and A Lempel. Unisys researcher Terry Welch later created an enhanced version of these methods, and Unisys holds a patent on the algorithm. It is widely used in many hardware and software products, including V.42bis modems, GIF and TIFF files and PostScript Level 2.
Mapping Placing an image on or around an object so that the image is like the object's skin.
Mask Information in the alpha channel of a graphic that determines how effects are rendered.
MDA Monochrome Display Adapter: the first IBM PC monochrome video display standard supporting 720x350 monochrome text but with no support for graphics or colours.
Mesh Model A graphical model with a mesh surface constructed from polygons. The polygons in a mesh are described by the graphics system as solid faces, rather than as hollow polygons, as is the case with wireframe models. Separate portions of mesh that make up the model are called polygon mesh and quadrilateral mesh.
Midtones Tones in an image that are in the middle of the tonal range, halfway between the lightest and the darkest tones.
Mip Mapping A sophisticated texturing technique to ensure that 3D objects gain detail smoothly when approaching or receding. This is typically produced in two ways; per-triangle (faster) or per pixel (more accurate).
Modelling The process of creating free-form 3-D objects.
Munsell Colour System A system consisting of over 3 million observations of what people perceive to be like differences in hue, chroma, and intensity. The participants chose the samples they perceived to have like differences.
NURBS Nonuniform Rational B-Spline: a type of spline that can represent more complex shapes than a Bezier spline.
On-The-Fly Switching A term used regarding the changing of resolution or refresh rates without having to restart a PC.
OpenGL Open Graphics Library: a standardised 2- and 3D graphics library that has its historical roots in the Silicon Graphics IrisGL library. It has become a de facto standard endorsed by many vendors and can be implemented as an extension to an operating system or a window system and is supported by most UNIX-based workstations, Windows and X Windows. Some implementations operate entirely in software, while others take advantage of specialised graphics hardware.
Particle Animation Rendering a 3D scene as millions of discrete particles rather than smooth, texture-mapped surfaces. Much more flexible but computer intensive.
PCX A popular bitmapped graphics file format originally developed by ZSOFT for its PC Paintbrush program. PCX handles monochrome, 2-bit, 4-bit, 8-bit and 24-bit colour and uses Run Length Encoding (RLE) to achieve compression ratios of approximately 1.1:1 to 1.5:1.
Perspective Correction Adjustment of texture maps on objects, viewed at an angle (typically large, flat objects) in order to retain the appearance of perspective.
Phong Shading A computation-intensive rendering technique that produces realistic highlights while smoothing edges between polygons.
Pixelisation Graininess in an image that results when the pixels are too big. Also referred to as Pixelated.
Polygon Any closed shape with four or more sides. In 3D, complex objects like teapots are decomposed, or "tessellated", into many primitive polygons to allow regular processing of the data, and hardware acceleration of that processing.
Polygon-Based Modelling Representing 3-D objects as a set or mesh of polygons.
Primitives Smallest units in the 3D database. Usually points, lines, and polygons representing basic geometric shapes, such as balls, cubes, cylinders, and donuts. Some 3D hardware and software schemes also employ curves, known as "splines".
QXGA Quad XGA: a QXGA display has 2048 horizontal pixels and 1536 vertical pixels giving a total display resolution of 3,145,728 individual pixels - 4 times the resolution of an XGA display.
Radiosity Complex methods of drawing 3D scenes, which result in photorealistic images. Essentially, they calculate the path that light rays follow from objects to the viewer, and all the accompanying reflections. Also known as ray tracing.
RAMDAC The RAMDAC converts the data in the frame buffer into the RGB signal required by the monitor.
Rasterisation Rasterisation is the conversion of a polygon 3D scene, stored in a frame buffer, into an image complete with textures, depth cues and lighting.
Refresh Rate Expressed in Hertz (Hz), in interlaced mode this is the number of fields written to the screen every second. In non-interlaced mode it is the number of frames (complete pictures) written to the screen every second. Higher frequencies reduce flicker, because they light the pixels more frequently, reducing the dimming that causes flicker. Also called vertical frequency.
Rendering Fundamentally this relates to the drawing of a real-world object as it actually appears. It often refers to the process of translating high-level database descriptions to bitmap images comprising a matrix of pixels or dots.
Resolution The number of pixels per unit of area. A display with a finer grid contains more pixels and has a higher resolution, capable of reproducing more detail in an image. More of an image can be seen, reducing the need for scrolling or panning.
Saturated Colours Strong, bright colours (particularly reds and oranges) which do not reproduce well on video; they tend to saturate the screen with colour or bleed around the edges, producing a garish, unclear image.
Saturation The colourfulness of an area judged in proportion to its brightness. For example, a fully saturated red would be a pure red. The less saturated, the more pastel the appearance. See also Chroma.
Scaling Process of uniformly changing the size of characters or graphics.
Setup The conversion of a set of instructions concerning the size, shape and position of polygons into a 3D scene ready for rasterisation.
Shading The process of creating pixel colours. Gouraud is a constant increment of colour from one pixel to the next, while Phong is much more complex and higher quality. Flat shading means no smooth blending of colours, each polygon being a single colour.
Specular Highlights A lighting characteristic that determines how light should reflect off an object. Specular highlights are typically white and can move around an object based on camera position.
Spline A 3D bezier curve used in modelling.
Spline-Based Modelling Representing 3-D objects as surfaces made up of mathematically derived curves (splines).
Sprite A small graphic drawn independently of the rest of the screen.
SVGA Super-VGA; when SVGA first came out it was used to describe graphics adapters capable of handling a resolution of 800x600 with support for 256 colours or 1024x768 with 16-colour support. It subsequently came to be used to indicate a capability of 800x600 or greater, regardless of the number of colours available.
SXGA Super XGA: a screen resolution of 1280x1024 pixels, regardless of the number of colours available.
Tessellation The process of dividing an object or surface into geometric primitives (triangles, quadrilaterals, or other polygons) for simplified processing and rendering.
Texel A textured picture element; the basic unit of measurement when dealing with texture-mapped 3D objects.
Texture A (2 dimensional) bitmap pasted onto objects or polygons, to add realism.
Texture Filtering Bilinear or trilinear filtering. Also known as sub-texel positioning. If a pixel is in between texels, the program colours the pixel with an average of the texels’ colours instead of assigning it the exact colour of one single texel. If this is not done, the texture gets very blocky up close as multiple pixels get the exact same texel colouring, while the texture shimmers at a distance because small position changes keep producing large texel changes.
Texture Mapping The application of a bitmap onto a 3D shape to give the impression of perspective and different surfaces. Texture maps can vary in size and detail, and can be ‘projected’ on to a shape in various different ways: cylindrically, spherically and so on.
Texture Memory Memory used to store or buffer textures to be mapped on to 3D polygon objects.
TIFF Tagged Image File Format: a popular file format for bitmapped graphics that stores the information defining graphical images in discrete blocks called tags. Each tag describes a particular attribute of the image, such as its width or height, the compression method used (if any), a textual description of the image, or offsets from the start of the file to "strips" containing pixel data. The TIFF format is generic enough to describe virtually any type of bitmap generated on any computer.
Time Line A scale measured in either frames or seconds; it provides an editable record of animation events in time and in sequence.
Transparency The quality of being able to see through a material. The terms transparency and translucency are often used synonymously; however, transparent would technically mean "seeing through clear glass," while translucent would mean "seeing through frosted glass."
Trichromatic The technical name for RGB representation of colour to create all the colours in the spectrum.
True Colour The ability to generate 16,777,216 colours (24-bit colour).
Tweening Also known as in-betweening; calculating the intermediate frames between two keyframes to simulate smooth motion.
UXGA Ultra XGA: a screen resolution of 1600x1200 pixels.
Vector Graphics Images defined by sets of straight lines, defined by the locations of the end points.
Vertex A dimensionless position in three- or four-dimensional space at which two or more lines (for instance, edges) intersect.
VESA Video Electronics Standards Association: an international non-profit organisation established in 1989 to set and support industry-wide interface standards designed for the PC, workstation, and other computing environments. The VESA Local Bus (VL-Bus) standard - introduced in 1992 and widely used before the advent of PCI - was a 32-bit local bus standard compatible with both ISA and EISA cards.
VGA Video Graphics Array: also referred to as Video Graphics Adapter. VGA quickly replaced earlier standards such as CGA (Colour Graphics Adapter) and EGA (Enhanced Graphics Adapter) and made the 640x480 display showing 16 colours the norm. Other manufacturers have since extended the VGA standard to support more pixels and colours. See also SVGA.
VGA Feature Connector A standard 26-pin plug for passing the VGA signal on to some other device, often a video overlay board. This feature connector cannot pass the high-resolution signal from the card and is limited to VGA.
Video Mapping A feature allowing the mapping of an AVI, MPEG movie or animation on to the surface of a 3D object.
Video Memory The graphics card RAM used in the frame buffer, the Z-buffer and, in some 3D graphics cards, texture memory. Common types include DRAM, EDO DRAM, VRAM and WRAM.
Video Scaling and Interpolation When scaled upwards, video clips tend to become pixelated, resulting in block image. Hardware scaling and interpolation routines smooth out these jagged artefacts to create a more realistic picture. Better interpolation routines work on both the X and Y axis to prevent stepping on curved and diagonal elements.
Virtual Desktop When a graphics card is capable of holding in its memory a resolution greater than that being displayed on the screen, the monitor can act as a ‘window’ onto the larger viewing area which may be panned across the "desktop".
VM Channel Vesa Media Channel, VESA's video bus which avoids the main system bus.
VRML Virtual Reality Modelling Language: a database description language applied to create 3D worlds. A standard for Internet-based 3D modelling, but still evolving.
VUMA VESA Unified Memory Architecture: a standard which establishes the electrical and logical interface between a system controller and an external VUMA device enabling them to share physical system memory.
Wireframe All 3D models are constructed from lines and vertices forming a dimensional map of the image. Then texture, shading or motion can be applied. Also referred to as Polygon Mesh.
WMF Windows Meta File: a vector graphics format used mostly for word processing clip art.
X Windows A windowing system developed at MIT, which runs under UNIX and all major operating systems. It uses a client-server protocol and lets users run applications on other computers in the network and view the output on their own screen.
XGA eXtended Graphics Array: also referred to as Extended Graphics Adapter. An IBM graphics standard introduced in 1990 that provides screen pixel resolution of 1024x768 in 256 colours or 640x480 in high (16-bit) colour. It subsequently came to be used to describe cards and monitors capable of resolutions up to 1024x768, regardless of the number of colours available.
XYZ Planes The three dimensions of space; each is designated by an axis. The x- and y-axes are the 2D co-ordinates, at right angles to each other. The z-axis adds the third dimension. Z-buffers accelerate the rendering of 3D scenes by tracking the depth position of objects and working out which are visible and which are hidden behind other objects.
YUV A colour encoding scheme for natural pictures in which luminance and chrominance are separate. The human eye is less sensitive to colour variations than to intensity variations. YUV allows the encoding of luminance (Y) information at full bandwidth and chrominance (UV) information at half bandwidth.