[Comp.Sci.Dept, Utrecht] Note from archiver<at>cs.uu.nl: This page is part of a big collection of Usenet postings, archived here for your convenience. For matters concerning the content of this page, please contact its author(s); use the source, if all else fails. For matters concerning the archive as a whole, please refer to the archive description or contact the archiver.

Subject: comp.sys.ibm.pc.hardware.video FAQ, Part 4/4

This article was archived around: 5 Dec 1997 01:31:18 GMT

All FAQs in Directory: pc-hardware-faq/video
All FAQs posted in: comp.sys.ibm.pc.hardware.video
Source: Usenet Version

Archive-name: pc-hardware-faq/video/part4 Posting-Frequency: monthly (second Monday) Last-modified: 1997/02/19 Version: 1.0 URL: http://www.heartlab.rri.uwo.ca/vidfaq/videofaq.html
********************************************************************** COMP.SYS.IBM.PC.HARDWARE.VIDEO Frequently Asked Questions - Part 4/4 ********************************************************************** Q) What is the pinout for a standard VGA/PGA/EGA/CGA connector? Standard 15 pin D-Sub VGA connector pinout ___________________________________________________ \ / \ 1 2 3 4 5 / \ / \ 6 7 8 9 10 / \ / \ 11 12 13 14 15 / \_____________________________________/ Pin # Description 1 Red Video 2 Green Video 3 Blue Video 4 Sense 2 (Monitor ID bit 2) 5 Self Test (TTL Ground) 6 Red Ground 7 Green Ground 8 Blue Ground 9 Key - reserved, no pin 10 Logic Ground (Sync Ground) 11 Sense 0 (Monitor ID bit 0) 12 Sense 1 (Monitor ID bit 1) 13 Horizontal Sync 14 Vertical Sync 15 Sense 3 - often not used Compaq (and perhaps some other companies) use the "Sense" lines as a way of telling what kind of monitor is connected. Newer monitors with DDC (also called Plug-n-play) use some of these pins. [From: Ashok Cates (acates@clark.net)] The ID bit pins in the 15 pin connector are shorted/left open to identify the type of monitor. I don't think they are very important anymore, as most cards have software to set resolutions, refresh rates etc. However, I think their functions are: ID bit 0 and ID bit 2 grounded: Dual frequency analog color interlaced (8514 or compatible) or variable frequency analog color interlaced. ID bit 0 grounded, ID bit 2 not connected: Fixed frequency analog color (8512, 8513, or compatible) or variable frequency analog color non-interlaced. ID bit 0 not connected, ID bit 2 grounded: Fixed frequency analog monochrome (8503 or compatible) or variable frequency analog monochrome. -ID bit 1 and ID bit 2 are usually connected together. -Monitor model numbers are for IBM monitors. Standard 9 pin D-Sub PGA/EGA/CGA connector pinout [From: Michael Scott (scott@bme.ri.ccf.org)] _______________________ \ / \ 1 2 3 4 5 / \ / \ 6 7 8 9 / \_____________/ IBM Adapters Pin Assignment CGA EGA PGA VGA TTL 16 colours TTL 16/64 col. Analogue Analogue 1 GND GND Red GND 2 GND Secondary Red Green GND 3 Red Primary Red Blue Red 4 Green Primary Green Composite Sync Green 5 Blue Primary Blue Mode Control Blue 6 Intensity Secondary Green Red GND GND /Intensity 7 not used Secondary Blue Green GND not used 8 H. Sync H. Sync Blue GND H. Sync /Comp. Sync 9 V. Sync V. Sync GND V. Sync Q) What are VGA/SVGA/UVGA/8514/a/XGA? The wonderful thing about PC's is that there are standards for so many different things. The problem is that every company has their own standards ;-). The lack of a widely accepted standard for >VGA pixel addressabilities is causing plenty of problems for manufacturers, system builders, programmers and end users. As a result, each vendor must provide specific drivers for each supported operating system for each of their cards. In the list above, VGA, 8514/a and XGA are standards established by IBM, and have been accepted to a greater (VGA), lesser (XGA) or even much less (8514/a) degree. The reason for this may be a backlash against IBM (due to royalty demands) or that video card vendors were not satisfied with the suggested standards. For a more detailed discussion of VGA, see 'What is VGA, and how does it work?' The 8514/a was the next graphics offering from IBM and provides three new video modes that are not available from the VGA controller. Computers with 8514/a hardware must also have a VGA controller, as the 8514/a does not support VGA video modes. The additional modes are: Type Pixel Max. # Colours Characters Addressability gfx 640x480 256 80x34 gfx 1024x768 256 85x38 (interlaced) gfx 1024x768 256 146x51 (interlaced) The 8514/a also has some smarts, as it is capable of performing video memory transfers, drawing lines and extracting rectangular areas of the display image. These are so-called accelerated features. The XGA has superseded the 8514/a. It was the first IBM display adapter to use VRAM, and can be configured with 500k or 1 Meg. Like the 8514/a, the XGA has accelerated features which make it faster than standard VGA for some operations. The new modes XGA introduced are: Mode Type Pixel Max. # Colours Characters Addressability 14 text 1056x400 16 132x25 - gfx 640x480 256/65535* - - gfx 1024x768 16/256* - *500k/1 Meg configurations SVGA & UVGA SVGA and UVGA are not established standards, and so their meanings vary depending on manufacturer. VESA VGA BIOS Extensions are the closest thing to an 'SVGA' standard. Most video cards currently available are called SVGA (Super VGA), which basically means that the card provides a superset of standard VGA calls and capabilities. This means that anything better than 640x400 and 16 colours is an SVGA mode. Some suggest that SVGA covers 800x600 modes, while UVGA (Ultimate VGA) refers to 1024x768. However, the absence of any real standard renders the term SVGA quite useless, and the term UVGA is not used frequently. The result of having no SVGA standard is that there are many (>10 !) different SVGA chipsets available, and none of them use a common programming interface. Many provide video acceleration capabilities, which free the system CPU to do other tasks, i.e. hardware cursor, BitBlt, etc. However, to use the SVGA video modes and advanced features, each chipset requires its own driver. This is why video drivers are required for Windows 3.1, Windows 95, OS/2 & XFree86. These drivers, combined with accelerated hardware, can provide enormous increases in video performance. If you are looking for a machine and would like SVGA capabilities, don't accept that a given video card or monitor is adequate just because it is advertised as supporting SVGA. Instead, decide what maximum pixel addressabilities and colour depths you want to use, and at what vertical refresh rates, and ensure that the models you are looking at provide those capabilities, and that software drivers are available for the operating systems and programs you will be using. Q) What is VESA SVGA? While some vendors of video hardware decided to provide support for 8514/a or XGA standards set by IBM, most defined their own 'SVGA' modes. As a result, no common programming interface was available which would allow generic SVGA code to be written. In order for programmers to be able to write code which would work on a wide range of 'SVGA' hardware, VESA (Video Electronics Standard Association) defined a standard interface for SVGA functions. It's more correct title is 'VESA VGA BIOS Extensions' and it incorporates functions which allow a program to determine what video modes (pixel addressabilities and number of colours) and other functions are available and how the video memory is accessed. Because many vendors already had proprietary extensions to the VGA standard implemented in their hardware, VESA VGA extensions use a software interrupt to access all of the programming routines. This means that a video card vendor can provide a VESA video driver (also called a TSR - Terminate and Stay Resident program) which can fill the role of inter- preter between VESA VGA compliant software and proprietary SVGA hardware. As a result, programmers can now write software that will work on a range of SVGA hardware, taking advantage of more colours and higher pixel addressabilities than are available with VGA. The video modes defined by VESA are: Mode # Pixel Colours Addressability 100h 640x400 256 101h 640x480 256 102h 800x600 16 103h 800x600 256 104h 1024x768 16 105h 1024x768 256 106h 1280x1024 16 107h 1280x1024 256 Nuts & Bolts Specifically, the VESA VGA extension provides information and hardware setup to the application program. It has six functions: Function 0 - Return Super VGA Information Function 1 - Return Super VGA mode information Function 2 - Set Super VGA video mode Function 3 - Return current video mode Function 4 - Save/Restore Super VGA video state Function 5 - CPU Video Memory Window Control These functions are all accessed by placing 4Fh in the AH CPU register, the desired function in the AL register, then generating an interrupt 10h. While this VESA standard doesn't define how 'accelerated' functions like hardware mouse cursors, BITBLT or typical GUI windowing operations should be accessed, it does provide a common set of instructions for determining information about and programming of higher pixel addressabilities and colour depths for video cards that have a superset of standard VGA functions. For more information, contact VESA at: Video Electronics Standard Association 2150 North First Street San Jose, CA 95131-2029 (408) 435-0333 (408) 435-8225 http://www.vesa.org/ Q) What should I consider in buying a video capture card? There are several factors that will determine which video card is the best for your purposes. It will depend on the number and type of video inputs, AD (Analog to Digital) conversion and system noise, frame rate, video overlays and whether video capture is to be integrated with other software. Grayscale and colour video capture cards are available. Grayscale cards are usually 8 bit, but some are available for 12 bit conversion. This means that the video intensity is sampled temporally, measured as a voltage, then divided into 2^8 (2^12) or 256 (4096) discrete levels. 8-bit provides enough gray levels for most applications and approaches the noise threshold in most video systems. Noise can be reduced in this or any colour system by frame averaging. Colour capture cards are available in 16, 24, 32 and more bit models. They convert the individual red, green and blue video streams into digital values separately, each stream being treated similarly to grayscale digitization. 16-bit cards discretize RGB into 5, 5 and 6 bits, and so can record 65535 different colours. 24-bit cards provide 8 bits for each pixel for a total of up to 16.7 million colours. 24- bit cards are also called Truecolour because most humans can distinguish 5-6 million colours. At 16.7 million, 24-bit colour can display more different colours than anyone can perceive. Cards that provide 32 bits or more of colour depth are usually Truecolour cards with overlay capabilities. The overlay planes (8 bits in the case of 32 bit) can be used to contain text or graphics overlays, or can store depth information (z-buffer). In addition, extra video memory can be used to double buffer the incoming digitized signal, up to doubling the frame capture rate. Video capture cards can often digitize different image sizes , though the most common is 640x480. 640x480 is the maximum image size that is meaningful for NTSC video signals. Keep in mind that while the horiz- ontal resolution of a television signal is quite high, the vertical resolution is limited to the number of scan lines displayed. A VCR provides ~250 lines, while S-video or laser disc provide over 300. Many video cameras provide more - closer to the 525 that the NTSC standard can handle. This means that the capture card has to integrate vertically (or subsample) to get 480 pixels vertically. This introduces a smoothing effect in the vertical direction and results in a less sharp picture. Capture cards are available which will digitize larger images, but they require special-purpose video equipment to be used to any advantage. Ensure that the resolutions you use maintain the screen aspect ratio. Various types of input signals can be digitized including NTSC, PAL, S-video and RGB. Some cards can handle all types, but most of the less expensive ones can only understand NTSC. Boards that can capture separate RGB signals can often be used to connect up to 3 grayscale video inputs. Many video cards come with simple frame capture programs, but if you are planning to integrate video capture with other operations on the computer, like collecting data from an AD card, adding text data as an overlay or changing video-in channels on-the-fly, you will have to do some programming. In this case you will need good programming libraries in a language you are familiar with for the video card. Some companies include libraries with their cards, but most charge extra. Most often libraries, when available, are for C or BASIC, and sometimes Pascal. Q) What type of camera do I need for video capture? The type and quality of camera you require depends on the application. In general, most home hobbyists will opt for an inexpensive one-chip CCD colour camera, while high-end video applications require a three- chip colour CCD or tube camera. Most scientific work requires the high definition grayscale of a monochrome CCD or tube camera. CCD vs. Tubes Charge Coupled Device (CCD) cameras are a solid-state, inexpensive and durable alternative. The same technology as is incorporated into camcorders is used in stand-alone CCD video cameras. CCD's consist of an array of light-sensitive material, which produces an electrical signal when struck with a light photon. As light photons continually stream through the lens and strike the CCD, they produce different voltages in corresponding CCD elements. By sampling the voltage generated at each element, an analog raster representation of light intensity is collected. This produces a grayscale representation of the sampled light image, where the maximum voltage corresponds to white, and the minimum corresponds to black. CCD's suffer from black noise (noise generated from an element even when no light photons are striking it) and relatively low light sensitivity, though newer CCD's are improving. CCD's have the advantage of low cost and high durability. Tube cameras use older tube technology instead of solid-state silicon. They are very light sensitive, and so are useful for low- light applications. In general, most tube cameras are used when CCD technology in inadequate. They are more expensive than CCD's and are more easily damaged by excessive light exposure. Colour CCD Two varieties of colour CCD's are available; one and three chip implementations. A three chip CCD uses three discrete CCD arrays, each with a colour filter in front of it: red, green or blue. Each CCD is sampled in a raster fashion, the same as for the grayscale device, above, and the result is a colour analog signal. Because they require three discrete CCD's, the three chip models are more expensive than one chip models and provide better colour reproduction. The latter use one CCD, and no colour filters. They consider the energy of the incoming photons, which determines their colour, to produce a red, green and blue value for each CCD element in the array. While cheaper, colour reproduction of one chip CCD's is inferior to three chip. Digital Cameras Although expensive and used less frequently than analog video cameras, digital cameras have the advantage of not requiring dedicated frame capture hardware in the computer. They are based on the same CCD technology as is described above. An example of a digital camera is the IndyCam which come with SGI Indy workstations. Also, hand-held portable digital cameras are available which can download images to your computer. Q) I want to add an MPEG card to my system. How does it work? The Motion Pictures Experts Group (MPEG) has released a series of standards which describe a lossy digital video compression technique. In some cases, MPEG can reach compression rates of 100:1. It works by removing redundant information and details that most people would generally miss, and in later versions storing only the differences between successive frames. When an MPEG video clip is viewed on the screen, the video stream must be decoded on-the-fly. If done in software, this operation can be quite demanding of the system CPU. An alternative is to have a dedicated coprocessor do the MPEG decoding, then feed the resulting video stream to the video card. Because this type of coprocessor is dedicated to MPEG decoding, it can be optimized to perform the operation very fast, and can also be used to scale up the size of the resulting video with little or no degradation in performance. Even a relatively small 320x200 video displayed at 30 frames per second requires a bandwidth of 15.4 million bits. This would seriously decrease available bandwidth for other purposes like disk i/o if all of that data was dumped down the peripheral bus (ISA, VLB, PCI, etc). As a result, many video card manufacturers incorporated a feature connector on their VGA cards. This connector gives direct access to video display memory, allowing high frame rate video to be dumped to the monitor. One limitation of this adapter is that it can only provide 8-bit (256 unique) colour. If you're planning on using your PC as a VCR, you'll be disappointed with an MPEG card playing the cdrom version of your favourite film. The resolution will be inferior to that provided by your television. If you want to get smoother video playback and/or free-up your CPU for other tasks, then the addition of an MPEG decoder card may be worth the cost. Q) What is the feature connector on my video card for? A The feature connector comes in two variants; VGA and VESA. The basic idea is that video memory can be directly accessed by using the feature connector, bypassing the CPU and peripheral bus. This reduces CPU load and avoids bandwidth bottlenecks. In addition, this eliminates the need for a separate RAMDAC. Typically, the feature connector is used by video capture cards or MPEG decoder cards, as it provides high bandwidth which is ideal for playing video clips. The older VGA feature connector is limited by the VGA itself, and can only display up to 256 colours at a pixel addressability up to 320x200. While this can provide reasonable images, they tend to look dithered, and the quality is less than that of NTSC television. A more recent standard is the VESA Media Channel (VMC) which allows you to attach an MPEG or TV tuner card to your video card thorugh a high-speed connector. The VMC actually implements a full bus system which allows up to 15 devices to share the frame buffer and RAMDAC on the video card. More information on the VMC is available at the VESA WWW site at www.vesa.org, though the standards themselves are only accessible to VESA members. Q) What is DCI? [From: Dylan Rhodes (Formerly of Hercules)] DCI stands for "Device Control Interface." It's an Intel/Microsoft standard, and exists primarily as a way for Windows 3.1 to exploit the video acceleration features of a graphics card, and/or to provide fast video when needed -- for example, the WinG games library uses DCI. A DCI driver exists at the same software layer as the GDI. Among DCI's capabilities are the ability to write directly to the frame buffer (helpful for high-speed games) and the ability to provide for on-board hardware acceleration of video scaling (i.e. stretching a video window to a larger size) and color space conversion (converting the YUV format color information in a video file to the RGB format that a typical graphics card RAMDAC expects). Note that support for DCI features doesn't need to be in hardware -- a graphics card vendor could provide a DCI driver that allowed Windows 3.1 apps to speak DCI, but the graphics card could be performing the DCI functions with a software driver. Note: with Windows 95, DCI will be replaced by an expanded interface called DirectDraw. Q) How do I contact my video card/monitor vendor? A large list of vendors' phone numbers is distributed in the c.s.i.p.h. FAQ section 9.8. This FAQ is posted monthly to c.s.i.p.h.* groups, and is available via FTP from: rtfm.mit.edu in /pub/usenet/comp.sys.ibm.pc.hardware.systems and its mirrors. Many vendors are on-line. See the following section "Is there an Internet ftp/web site for my video card?" and/or these sites for listings: http://www.ronin.com:80/SBA/ : Guide to Computer Vendors by SBA Consulting. http://mtmis1.mis.semi.harris.com/comp_ph1.html#top : Computer Companies Phone List Pt 1 - HARRIS Mountaintop http://www.cviog.uga.edu/monitors/manufacturers.html : List of 60+ monitor companies with phone numbers and WWW sites. Q) I need new drivers. Is there an Internet ftp/web site for my video card? It's pointless for me to try to maintain a list of on-line services, since they change so frequently, and others are already doing it! Please refer to: comp.sys.ibm.pc.hardware FAQ in sections 9.6 and 9.7. This FAQ is posted monthly to c.s.i.p.h.* groups, and is available via FTP from: rtfm.mit.edu in /pub/usenet/comp.sys.ibm.pc.hardware.systems and its mirrors. http://www.rust.net/~frankc/ : List of Windows95 drivers and updates Here's a list of ftp sites for video card vendors on the 'net for those too lazy to look up the other sources. For a larger list of video related WWW sites including vendors and information, refer to: http://www.heartlab.rri.uwo.ca/vidfaq/vendors.html Most major vendors' www sites are listed there. FTP Sites: Alliance Semiconductor Corp. ftp://www.alsc.com ATI Technologies Inc. ftp://ftp.atitech.ca Boca Research Inc. ftp://ftp.bocaresearch.com Cirrus Logic ftp://ftp.cirrus.com Diamond Multimedia Systems, Inc ftp://ftp.diamondmm.com ELSA ftp://www.elsa.com/ Genoa Systems Corp. ftp://www.genoasys.com/ Hercules ftp://ftp.hercules.com Matrox ftp://ftp.matrox.com Media Vision ftp://ftp.mediavis.com/ MIRAGE Video Solutions ftp://ftp.mirage-mmc.com/ miro Computer Products ftp://ftp.miro.com Number Nine Visual Technology ftp://ftp.nine.com Radius ftp://ftp.radius.com S3, Inc. ftp://ftp.s3.com Software Integrators ftp://ftp.avicom.net/pub/softint STB Systems, Inc. ftp://ftp.stb.com/ UMAX ftp://www.umax.com/ VideoLogic ftp://ftp.videologic.com/ If you have any more sites to add, please email the FAQ maintainer. In addition, video drivers are archived at: ftp://ftp.cdrom.com windows: /.22/cica/drivers/video linux: /.6/linux OS/2: /.4/os2/drivers & maybe /.4/os2/warp ftp://ftp.microsoft.com/ http://www.jumbo.com/ Other video utilities are available at SimTel, Cica and Garbo mirrors. ********************************************************************** References ********************************************************************** Abrash, Michael. Demystifying 16-bit VGA, in Dr. Dobb's Journal, May 1990 Abrash, Michael. Mode X: 256-color VGA magic, in Dr. Dobb's Journal, July 1991 Ericsson, Bo. VESA VGA BIOS extensions, in Dr. Dobb's Journal, April 1990 Howard, Christopher A. Super VGA programming, in Dr. Dobb's Journal, July 1990 McNierney, Ed. New issues in PC graphics, Dr. Dobb's Journal, November 1986 Myers, Ben. Saving and restoring VGA screens, in Dr. Dobb's Journal, July 1991 Norton, P. Inside the IBM PC and PS/2, 4th Edition. Brady, New York, New York, c1991 Peddie, Jon. High-resolution graphics display systems. Windcrest (McGraw-Hill), U.S.A., c1994 Sanchez, J. Graphics design & animation on the IBM microcomputer, Prentice Hall, Englewood Cliffs, N.J., c1990 Sutty, G. & Blair, S. Advanced programmer's guide to the EGA/VGA, Brady, New York, New York, c1988 ********************************************************************** Acknowledgments ********************************************************************** Contributing Authors: Ron Bean (rbean@execpc.com) Sam Goldwasser (sam@stdavids.picker.com) Declan Hughes (hughes@cat.rpi.edu) Bill Nott (BNott@bangate.compaq.com) Dylan Rhodes (Formerly of Hercules) Michael Scott (scott@bme.ri.ccf.org) Roger Squires (rsquires@cyclops.eece.unm.edu) Ralph Valentino (ralf@alum.wpi.edu) Reviewers: Sam Goldwasser (sam@stdavids.picker.com) Andy Laberge (tic-toc@wolfe.net) Bill Nott (BNott@bangate.compaq.com) Dylan Rhodes (Formerly of Hercules) Ralph Valentino (ralf@alum.wpi.edu) ********************************************************************** Appendix A - Glossary ********************************************************************** ********************************************************************** GLOSSARY OF TERMS FOR THE VIDEO FAQ ********************************************************************** If you don't find the definition you are looking for in this glossary, try the resources below: The "Free On-line Dictionary of Computing" is available via the web at: http://wombat.doc.ic.ac.uk/ This dictionary is compiled and maintained by Denis Howe (dbh@doc.ic.ac.uk). A large list of COMPUTER ACRONYMS is defined in the Babel document accessible via ftp or the web. It is updated 3 times per year, so you have to request the latest document. It's of the format babelYRP.html where YR is the year, i.e. 95 P is the update period i.e. a, b or c: After May 1, 1995 request BABEL95B. After Sep 1, 1995 request BABEL95C. After Jan 1, 1996 request BABEL96A. http://www.access.digex.net/~ikind/babel95b.html ftp://ftp.temple.edu/pub/info/help-net filename as above babelYRP.txt i.e. babel95b.txt Babel is compiled and maintained by Irving Kind (ikind@mcimail.com). ********************************************************************** Glossary ********************************************************************** 8514/a IBM video graphics standard. Supports pixel addressabilities up to 1024x768 and 256 colours. It is _not_ a superset of VGA. addressability (pixel addressability) This refers to the number of pixels that a video controller can display. It is quoted as the (# horizontal pixels) by the (# vertical pixels). Common PC pixel addressabilities include: 320x200, 640x480, 800x600, 1024x768, 1280x1024 & 1600x1200 aperture grille An array of vertical wires which act in a similar manner as a shadow mask. Their basic purpose is to permit the correct electron beam to strike its corresponding colour phosphor only. This results in crisp pixel definition, and superior colour brightness than is realized with more traditional designs. The aperture grille was first used by Sony in their Trinitron design. AT bus Advanced Technology (IBM) bus. The standard PC compatible peripheral bus to which add-in cards like video, i/o, internal modems, sound are added. Also called the ISA bus, it runs at a maximum of 8.33 MHz and has a 16-bit wide data path. bandwidth Also called video bandwidth. This is a measure of how much gross throughput a monitor can handle (in MHz). Bandwidth at a given pixel addressability is a function of the vertical refresh rate and monitor timing. see 'How do I calculate the minimum bandwidth required for a monitor?" BIOS Basic Input Output System. The video BIOS basically tells the computer how to talk to the video subsystem at boot time. The video BIOS calls are used by DOS for VGA (and SVGA) modes. BITBLT A VGA video operation which copies an array of values to a rectangular region in video RAM. bit planes This is the number of bits which are available to store colour information for each pixel displayed. The number of colours which can be displayed is calculated as two to the exponent 'n', where n is the number of bit planes. i.e. 4 bit equals 16 colours, 8 bit equals 256 colours and 24 bit equals 16.7 million colours. see "How does colour depth (bit planes) relate to the number of colours?" colour depth Refers to the amount of memory (and therefore number of simultaneously displayable colours) available to store colour information for each pixel. see 'bit planes'. CPU Central Processing Unit. This is the heart and brains of your computer. It is responsible for executing code, moving data, calculations, etc. For PC's, this chip is a member of the X86 family including 8088 through 80486, Pentium and Nextgen. CRT Cathode Ray Tube. Basically the same technology as is in modern television sets. One or more beams of electrons are focused onto phosphor, causing it to glow. The phosphor is arranged into an array (usually close to rectilinear), and the electron beam scans the phosphor on the screen (similar to how you read text - left to right and top to bottom), usually 60+ times per second. degauss Magnetic interference caused by a change in the position of a monitor in relation to the earth's magnetic field or the presence of an artificial magnetic field can cause discolour- ation. To correct this, all colour monitors automatically degauss at power-on and some also have a manual degaussing button. This allows the monitor to compensate for the change in the magnetic field by realigning the electron guns. In some low cost monitors without degauss buttons it is necessary to leave the power turned off for at least 20 minutes in order to get maximum degaussing. display Usually used to indicate the monitor or flat-panel device used as the primary visual interface. display adapter Usually this is the same as the video card, but some mother- boards have built-in video, and so don't require an additional card. The display adapter contains video memory which stores what is displayed on the computer's monitor. They have a wide range of features, from a basic frame buffer, to advanced 3D geometric rendering engines. dot clock Technically, this refers to the digital clock signal that transfers data into the video card's digital to analog converter. However, it has also become a measure of the maximum gross data throughput of a monitor. It is measured in MHz, and indirectly determines the maximum pixel addressability and vertical refresh rate that a monitor can handle. See "What do those monitor specifications mean?" dot pitch The distance between a phosphor dot of one phosphor triad to its closest diagonal neighbour of the same colour on a monitor. Expressed in mm - i.e. .28 dot pitch means .28 mm between triads. A smaller value indicates that the phosphor dots are more closely spaced, and that the resulting image displayed will be crisper. dot stripe see 'stripe pitch' DRAM Dynamic Random Access Memory. The vast majority of system RAM in modern computers is of this type because of it's low cost. It is also the most common type of RAM used for video cards. A specialized type of DRAM called VRAM is also used in higher end video cards. see "What is the difference between VRAM and DRAM?" EGA Enhanced Graphics Adapter (IBM). Precursor to VGA, all EGA video modes are supported in VGA, though register compatibility is not 100%. EGA cards generate a digital signal, and thus will not drive a modern, analog monitor. EISA Extended Industry Standard Architecture. This 32-bit bus standard was created primarily to compete with IBM's MCA bus. It runs at speeds of up to 8.33 MHz. EISA is a dying standard. graphics controller This is a generic term to describe the video hardware in a computer. Sometimes it is built onto the motherboard, but usually it is a separate daughter card that fits into one of the expansion bus slots. The interface between the graphics controller and the main processor is one of the ISA, EISA, MCA, VLB or PCI buses. The graphics controller is responsible for generating the video signal that is sent to the monitor. Typically a graphics controller contains a graphics coprocessor which may be a graphics accelerator, video RAM and a RAMDAC. graphics coprocessor A secondary processor dedicated to performing video display tasks. graphics accelerator This is a highly misused and now almost meaningless term. For the purposes of this FAQ, a graphics accelerator is a coprocessor which is capable of specific graphics operation, independent of the main system CPU. See the section "How does a video accelerator work, and will one help me?" GUI Graphical User Interface. In contrast to text-based interfaces like DOS or UNIX, GUI's provide more flexibility in terms of colour, pixel addressability and types of objects that can be displayed. Examples of GUI's include X-Windows, Microsoft Windows 3.1, OS/2. Hercules A monochrome display adapter which is MDA compatible and provides graphics modes up to 720x348 horizontal refresh see horizontal scan rate horizontal scan rate (horizontal frequency) The frequency, expressed in kHz (thousands of times per second), at which the horizontal deflection circuit operates. This roughly translates to the number of scanlines displayed on a monitor in one second. interlaced Standard NTSC television signals are interlaced, meaning that each video frame is divided into two separate fields of alternating scanlines. The resulting fields are displayed sequentially, such that what was originally a 30 frame per second (fps) refresh becomes 60 Hz at half the vertical pixel addressability. Thin horizontal lines will appear to flicker on an interlaced display since their effective refresh rate is only 30 Hz. ISA Industry Standard Architecture. This is a 16-bit bus standard which runs at speeds of up to 8.33 MHz. The vast majority of peripheral add-in cards like modems, sound cards, cdrom interfaces and other low-bandwidth applications are still ISA based. VLB and PCI provide higher bandwidth for video and disk I/O operations. Look-up Table (LUT) At higher pixel addressabilites, most graphics controllers can not simultaneously display as many colours as they are capable of generating. Because of video card memory limitations, only a subset of all possible colours can be displayed at one time. A look-up table stores the mapping information which determines which subset of all possible colours are available at any given time. MDA Monochrome Display Adapter (IBM) monitor Usually a CRT-based device which directs an electron beam onto coloured phosphor, causing it to glow. Monitors use the same basic technology as televisions, but are capable of much higher pixel addressabilities and resolutions. motherboard The main component of the computer, which contains the CPU (brain), main memory slots, keyboard connector and expansion bus slots, among other possible components. non-interlaced This means that an entire frame is displayed with each screen refresh. Non-interlaced displays produce a more pleasing screen image since thin horizontal lines don't flicker with each screen refresh. OEM Original Equipment Manufacturer. Often manufacturers will produce versions of their products in large quantities for other companies who either stick their name on them or use them as components for their systems. OEM products often make it to the retail sales arena where they are sold at lower prices. An OEM version of a card _may not_ be equivalent to the retail version. PCI Peripheral Components Interconnect. This is basically the Pentium equivalent to the VLB, but with improvements. It is a 64-bit standard, but is currently only implemented as 32 bits - look for 64 bit PCI in the future. It performs asynchronously to the main CPU, meaning that the PCI bus operates at 33 MHz regardless of the CPU clock. It also allows more than two devices on the bus, unlike VLB. phosphor triad (dot triad) This is the smallest dot that can theoretically be resolved on a colour monitor and consists of three phosphor dots - one each of red, green and blue. When struck with the electron beam, these dots glow producing a bright spot on the screen. Practically, 1.2 or more dot triads comprise each pixel on the screen, although the pixel addressability of some monitors is greater than their resolution, and in this case a pixel can be smaller than a dot triad. The result in this case is that small objects may not be resolvable. pixel This is the smallest addressable display unit available at a given video addressability. There is no physical thing on a display that can be called a pixel. Pixels exist only in the graphics controller bitmap. The screen image in the bitmap is composed of an array of pixels, arranged in a rectilinear fashion, with the X axis running horizontally, perpendicular to the Y axis. A pixel consists of intensity only (in grayscale monitors) or colour and intensity information (red, green & blue in colour). While a pixel usually corresponds to a square or rectangular area, it is displayed as a number of spots on a CRT. One pixel usually consists of 1.2 or more dot triads. Flat panel displays are a special case where individual pixels correspond directly to a picture element on the display. pixel addressability see addressability RAM Random Access Memory. RAM comes in different types, including DRAM (Dynamic RAM) and VRAM (Video RAM) among others. DRAM is used as main system memory, while both DRAM and VRAM can be used on graphics cards. RAMDAC Random Access Memory Digital-to-Analog Converter. This is part of the graphics card which converts the digital intensity values for each of the red, green and blue guns (usually an 8-bit number) to analog voltages which are sent to the monitor. A RAMDAC can use its RAM to store look-up table (LUT) information. refresh rate When referring to monitors, the number of times that the video card refreshes the entire screen in one second. Expressed in Hz (Hertz). resolution The most common misinterpretation of this term is that it is the same as pixel addressability. In fact, resolution is more closely related to dot pitch, since it is a limitation of the monitor rather than of the graphics controller. The resolution limits how small an object a monitor is able to display. RGB Red, Green and Blue. By varying the intensity of each of these colours in a single pixel, the human eye can be fooled into seeing a wide range of colours. For example, a combination of red and green appears as yellow, even though no light with a yellow wavelength is emanating from the screen. This works because the optical system integrates the photons striking a region on the retina, and the combined impulses from green and red sensitive cones are seen as yellow. scanline The movement of a monitor's electron gun from one side of the screen to the other results in the appearance of a horizontal line of varying intensity and colour. Typically, 200 to 1200 horizontal scan lines (lined-up vertically on top of each other) make-up the image you see on your display. shadow mask This is usually an invar mask which acts to block the electron beam from striking the wrong phosphors in a CRT. The beam passes through holes in the mask to strike the correct phosphor while shadowing neighbouring phosphor. i.e. it prevents a beam intended to strike a red phosphor from striking a neighbouring green phosphor by causing an electron shadow over the green dot. stripe pitch This is similar to dot pitch, but applicable to Sony Trinitron and similar tubes which use fine vertical wires (aperture grille) to separate phosphors. Dot stripe is measured as the distance between the vertical stripes that result. Measures of dot pitch and dot stripe are _not_ directly comparable. Trinitron A common but proprietary picture tube design developed by Sony. Uses fine vertical wires instead of the more traditional shadow mask. see "Why does my monitor have 1/2/3 faint horizontal lines on it?" vertical refresh rate (vertical scan rate) The number of fields (on an interlaced display) or frames (on a non-interlaced display) that are displayed in one second. A field or frame covers the entire screen area. This is measured in Hz (cycles per second). It is limited by the monitor and video card (pixel addressabilities and colour depths). Modern monitors and video cards provide refresh rates of 60Hz+. VESA Video Electronics Standards Association. This group has produced standards for the VLB (Vesa Local Bus), VESA SVGA video modes and standards for minimum screen refresh rates at various pixel addressabilities. VGA Video Graphics Array (IBM). Supports pixel addressabilities of up to 640x480x16. This is the de facto video standard and consists of a number of video modes. It is still heavily supported by DOS-based applications and games. see "What is VGA, and how does it work?" video card A dedicated piece of hardware which performs graphics operations. Also called a display adapter. Consists of microchips and other electronic components mounted on a pc-board which connects into a slot (ISA, EISA, MCA, VLB or PCI) on the motherboard. viewable area Typically monitors are advertised by the diagonal size of the picture tube in inches. Common sizes are 14", 15", 17", 20"+. However, the amount of the screen that can be seen is usually less. For example, most 17" monitors have only a 15.5" diagonal area used for display, in part because the actual phosphor area is only about 16" due to the glass thickness. This is partially due to the fact that the monitor's case covers the edge of the tube, and partially because monitor manufacturers want to make you think you're getting a larger display than you are. see "Size" under "What do those monitor specifications mean?" VLB VESA Local Bus. This 32 bit bus was originally designed to provide higher bandwidth for video cards than is available with the ISA bus. It is optimized for the 486 CPU and can run at speeds up to 40 MHz with one card on the bus, or up to 33 MHz with two cards on the bus. The speed of the VLB is dependent, and runs synchronously with, the main system CPU. Some VLB cards are not designed to run faster than 33 MHz, though some mother- boards will clock the bus at up to 50 MHz! VLB 2.0 has been written, but has not been implemented on many 486 motherboards. VRAM Video Random Access Memory. A specialized type of DRAM, VRAM is dual-ported, meaning it can be read from and written to at the same time. see "What is the difference between VRAM and DRAM?" ********************************************************************** Appendix B - Popular Video Chipsets ********************************************************************** Some of the information in this section was taken from Boogyman's video FAQ. For more detailed chipset info, or info on a particular video card model, refer to the Chipsets document which is distributed as a supplement to this FAQ. The relevant sections are included with the Video Chipset Information List. Note that while much of the information in this section is fact, by its very nature, some of the following is based on opinion. Please don't submit flames - any messages that are of the 'my video card is the best' will be sent to /dev/null. If you disagree with any of the comments below and can provide reasonable justification, feel free to email the FAQ maintainer. Note that 'my card is _really_ fast' isn't good enough evidence. Alliance Promotion 32, 64 and 128 bit chipset with DRAM, EDO and video acceleration capabilities. Their 6422 chipset is faster than the CL5434 and some S3 chips for Windows (according to Alliance) but DOS/VGA speed is poor. ARK 1000PV, 2000PV A relative newcomer to the video arena, the ARK2000 based cards are the fastest DOS/VGA performers available. They are inexpensive, 32 (1000) & 64 bit (2000) processors that provide some acceleration capabilities, but are only available in DRAM versions. Of course, for any VGA application, VRAM would provide no benefit over DRAM anyway. (i.e. Hercules Stingray 64/Video, Actix Picasso 64) ATI (see Mach 32 & Mach 64) Avance Logic Inc. -ALI (ALG2301/ALG2228) 32 bit DRAM based chipset with limited acceleration features. A 64-bit chip is also available. Cirrus Logic (542x) C.L. based cards have become the de facto entry level video adapters for VLB and PCI. While somewhat faster than the older Trident 8900, C.L. chipsets are economy models, low on price and acceleration. These chipsets are 32 bit and support DRAM up to 2 MB for some models. Because they are so common, they are well supported across different operating systems. Cirrus Logic (543x/544x) This is the 64 bit replacement for the 542x series of chips. Has some acceleration features and in general are good cards for their modest price. The GD5430 has only 32 bit DRAM interface even with 2 MB installed, but other models have full 64 bit interface. Mach32 While no longer in production, this is a popular 32 bit chip. Its speed has been surpassed by many newer chipsets, but it still provides good performance, and is supported widely. Supports up to 2 MB of VRAM or DRAM and 64 bit memory transfers (interleaved). (i.e. ATI Graphics Ultra Pro) Mach64 (88800) This 64 bit chip was designed by/for ATI. It provides accelerated GUI performance and respectable VGA speed. Support across many OS's and most buses is available. Mach64 based cards have been given first place honours in many PC magazine video card rankings due to good performance and excellent drivers and utilities. Newer versions of the Mach64 provide video acceleration, while the Rage 3D accelerator adds 3D acceleration. (i.e. ATI Graphics Pro Turbo, WinTurbo, Xpression, Video Xpression) Matrox MGA Typically, Matrox cards are blisteringly fast for GUI's (typically Windows 3.1) and are considered a high-end chipset because of their cost. However, the VGA chipset used on Matrox cards is abysmal, and is usually much slower than even the cheapest VGA cards. Matrox has released a new card called the Millenium which is very fast for GUI's and has fast VGA performance. The newer Mystique also provides fast GUI and DOS/VGA speed. It uses the new high-bandwidth, low-cost WRAM technology. Oak Technologies Inc. Provide low end SVGA chipsets, some available with up to 2 MB. Performance isn't spectacular, and are fairly inexpensive. S3 ViRGE, Vision864, Vision868, Vision968, Trio64, 805, 911, etc One of the most popular chipsets, S3-based cards seem to appear at or near the top of most Windows 3.1 accelerator top ten rankings. The S3 family enjoys good support across most operating systems and the 864/964 and Trio64 provide very fast performance for GUI's and respectable VGA speed. Typically the 8xx series are DRAM based, while the 9xx are for VRAM. The 911 and 924 were the first generation chips which came with VRAM. Provide good GUI acceleration but poor VGA performance. The 32 bit chipsets consist of 801, 805, 928 and Trio32 (732). The 801 is a low end chip which is faster than comparable C.L. chips. The 805 supports VLB and the 805i supports interleaved DRAM. The 928 is a high-end 32 bit card and the Trio32 is an attempt to dominate the low-end 32 bit market - used DRAM. The 64 bit chipsets are ViRGE, 864, 964, Trio64 (764),V+ 868 and 968. These chipsets dominate many top 10 listings for economical but fast GUI accelerators. The Trio64 is basically an 864 with integrated RAMDAC but isn't software compatible. All have a 64 bit memory interface. The x68 chips have additional video acceleration capabilities, as does the Trio64V+. The ViRGE has a Trio64V+ core with additional 3D acceleration features. (i.e. Hercules Terminator 64/DRAM (Trio64), STB Powergraph 64, Diamond Stealth 64 (Trio64), rPC FireStorm64, Paradise Bahamas 64 (Vision864), Diamond Stealth Video (Vision868), Hercules Terminator Professional (Vision868). Trident For a long time, Trident chipsets (89xx) were the most common entry level chips used on ISA boards. They aren't very fast for VGA or GUI but enjoy broad-based SVGA mode support. The newer Trident chips (94xx) are faster and provide some acceleration features, but are still one of the slowest chipsets available. The more recent 9680 is a respectable GUI accelerator. TSENG ET6000 Tseng's latest chipset gives blazing DOS/VGA performance and fast Windows performance. Many vendors offer ET6000 based cards. It supports DRAM and EDO DRAM sporting a 128-bit accelerator and on-chip DAC. ET4000 - W32, W32i, W32p Traditionally the fastest VGA chipset, the ET4000 has recently been put into second place behind the ARK2000 chipset(s). The original ET4000 chipset is not accelerated, but the newer W32's are. It enjoys good support across most OS's, but lacks the bandwidth (being a 32 bit chip) to provide high-resolution, high-refresh truecolour performance. The 'p' denotes PCI, 'i' and 'p' are capable of memory interleaving. (i.e. Hercules Dynamite Pro(W32i) or Power (W32p), Cardex Challenger, Diamond Stealth 32 (discontinued)) UMC 8710 This is a less common 32-bit interleaved DRAM-based chipset. Weitek P90xx and P91xx Very fast VRAM-based accelerators have no VGA support which must be provided by a separate chip. 32-bit. Weitek is not providing drivers for Windows95, so support is vendor-specific and spotty at best. Western Digital (Paradise) The WD90C3x chipsets are 32-bit DRAM based and are available in up to 2 MB version. Have some acceleration capabilities. Philips Electronics purchased Western Digital's video chipset division in 1995. ********************************************************************** Appendix C - Circuit for Converting from VGA to Fixed-Freq. RGB ********************************************************************** More information and additional circuits are available on the PC Video FAQ WWW Site. They have not been included here for brevity and because some of the information is presented in a graphical manner. Please refer to http://www.heartlab.rri.uwo.ca/vidfaq/fixed.frequency.html for more information and superior circuits for driving workstation monitors. The following is an ASCII file of a circuit which combines the vertical and horizontal sync signals from a VGA card to output a composite sync signal compatible with many fixed frequency monitors. If necessary, the Csync line can be connected with the green video line to produce a sync-on-green signal. This circuit produces a signal that is compatible with many fixed-frequency monitors, but ensure that the signal you are sending has the same vertical and horizontal frequencies as the monitor expects. For more information on how others have done this, read pertinent parts of this FAQ and refer to: http://rugmd0.chem.rug.nl/~everdij/hitachi.html - Experiences hooking up a Hitachi monitor http://www.midcoast.com/jp/sun/ - More experiences with a Hitachi http://www.devo.com/video - Fixed Frequency PC Video FAQ *** USE AT YOUR OWN RISK - Others have used this circuit but I have not! [From: Roger Wolff (R.E.Wolff@et.tudelft.nl)] VGA connector monitor R ----------------------------------------------------- R gnd ----------------------------------------------------- gnd G ----------------------------------------------------- G gnd ----------------------------------------------------- gnd B ----------------------------------------------------- B gnd ----------------------------------------------------- gnd ___________ | | | 74HCT86 | | | | | hsync -----------|1 | | 3|----------------------------- Csync vsync -----------|2 | | | | | | | gnd--+--|7 14|--+-- +5V | |___________| | | | |________||_______| || 0.1 uF Use COAX cables for the "data" (R, G, B) lines. You can use just about anything for the hsync and the vscyn lines but keep them as short as possible. Around 30 cm (a foot) is fine. I use a COAX cable for the Csync line too, as I need a BNC connector at the end of the monitor anyway. I didn't do anything about termination, and all seems to be A-OK. I gather the power for the 'HCT86 from inside the computer. You can find something yourself (find a 5V powerline going to the leds or something), or use something that resembles those "add-on" fans. The latter usually use 12V, but they show the principle: a male and a female cable connector and 6 wires should do the trick. This worked for my Grayscale monitor (where the R and B lines are not needed), and now works just fine for my 21" monochrome monitor. I start X at boot-time, before ANYTHING runs: my Xserver owns PID 3. I then quickly bring up the network and start an Xconsole. This allows me to follow the rest of the bootsequence. ********************************************************************** END of comp.sys.ibm.pc.hardware.video FAQ - Part 4/4 ********************************************************************** -- Michael J. Scott R.R.I., U of Western Ontario mjscott@heartlab.rri.uwo.ca 'Need a good valve job?' PC Video Hardware FAQ: http://www.heartlab.rri.uwo.ca/videofaq.html ############### Illegitimus non tatum carborundum. ##############