1.3 Graphics Workstation Graphics workstation is the Graphics Kernel System's (GKS) term for a graphical device that can display graphical output, or accept graphical input, or both. The Fig. 1.5 shows the block diagram of typical graphics station. It consists of central processing unit (CPU), plotter, joy stick, keyboard, mouse, light pen, scanner etc, The main hardware components of a graphics workstation are CPU and display processor. The display processor is also called a graphics controller or a display coprocessor. It makes CPU free from the graphics chores. In addition to the system memory, a separate display processor memory area is provided in graphics workstations. Graphics workstations splay processor, memory, display devices, recorder, have a provision to interface video cameras and television ‘set. Graphics workstations also have display devices with very high resolutions and very high performanceComputer Graphics 1 Introduction to Computer Graphics CPUs. The size of the display device, colour supported by it, whether it is a raster or line drawing device are the main Properties of graphics. workstation. Manytimes, the fee, Or ie) connected to local ‘arse te for sharing files, accessing shared Peripherals and other high performance computers, The graphics workstation is always Supported with graphics software. Graphics Software acts.as a very powerful. tool to create Scenes, images, pictures and also animated pictures. % Local area network interface Local Area Network Fig. 1.5 Block diagram of typical workstationComputer Graphics 1-8 Introduction to Computer Graphics 1.4 Graphics Hardware Graphics hardware includes the display system, input devices and display devices. In this section we see the details of display systems. 1.4.1 Architecture of a Vector Display As shown in Fig. 1.6 vector scan CRT display directly traces out only the desired lines. on CRT ie. If we want a line connecting point A with point B on the A vector graphics display, we simply drive the beam deflection circuitry, which will cause beam to go directly from point A to B. If we want to move the beam from point A to point B without showing a line between Fig. 1.6 Vector scan CRT points, we can blank the beam as we move it. To move the beam across the CRT, the information about both, magnitude and direction is required. This information is generated with the help of vector graphics generator. The Fig. 1.7 shows the typical vector display architecture. It consists of display controller, Central Processing Unit (CPU), display buffer memory and a CRT. A display controller is connected as an I/O peripheral to the central processing unit Fig. 1.7 Architecture of a vector displayComputer Graphics 1-9 Introduction to Computer Graphics (CPU). The display buffer memory Stores the computer produced display list or display program. The program contains Point and line plotting commands with (x, y) or Gy, z) end point co-ordinates, as well as character plotting commands. The display controller interprets commands for Plotting points, lines and characters and sends digital and point co-ordinates to a vector Senerator. The vector generator then converts the digital co-ordinate values to analog voltages for beam-deflection circuits that displace an electron beam writing on the CRT's phosphor coating. In vector displays beam is deflected from end point to end point, hence this technique is also called random scan. We know as beam, strikes phosphor it emits light. But phosphor light decays after few milliseconds and therefore it is necessary to repeat through the display list to refresh the Phosphor at least 30 times per second to avoid flicker. As display buffer is used to store display list and it is used for refreshing, the display buffer memory is also called refresh buffer 1.4.2 Architecture of a Raster Display In early seventies the inexpensive raster graphics displays were developed, Raster displays store the display primities (such as lines, characters solid and filled patterns) in a refresh buffer in terms of their corresponding pixels, cPU 90000000000000000000 100000000000000000000 1200000000000000000000 10000044 Refresh buffer Fig. 1.8 Architecture of a raster displayComputer Graphics 4-10 Introduction to Computer Graphics The Fig. 1.8 shows the architecture of a raster display. It consists of display controller, central processing unit (CPU), video controller, refresh buffer, keyboard, mouse and the CRT. ‘As shown in the Fig. 1.8, the display image is stored in the form of 1s and 0s in the refresh buffer. The video controller reads this refresh buffer and produces the actual image on the screen. It does this by scanning one scan line at a time, from top to bottom and then back to the top, as shown in the Fig. 1.8. The raster scan display system is the most common method of displaying images on the CRT screen. In this method, the horizontal and vertical deflection signals are generated to move the beam all over the screen in a pattern shown in the Fig. 1.9. Vertical Horizontal retrace retrace Here, the beam is swept back and forth from left to right across the screen. When the beam is moved from the left to right, it is ON. The beam is OFF, when it is moved from right to left as shown by dotted line in Fig. 1.9. When the beam reaches the bottom of the screen, it is made OFF and rapidly retraced back to the top to start again. This is known as vertical retrace. The time required to move electron beam from right bottom of the screen to left top of the screen is known as vertical retrace time. A display produced in this way is called raster scan display. Raster scanning, process is similar to reading different lines on the page of a book. After completion of scanning of one line, the electron beam flies back to the start of next line and process repeats. The movement electron beam from end of scan line to the start of next scan line is known as horizontal retrace. The time. required to move electron beam from end of scan line to the start of next scan line is. known as horizontal retrace time. In the raster scan display, the screen image is. maintained! by repeatating scanning the same image. This process is known as. refreshing of screen. ) Fig. 1.9 Raster scan CRT In raster scan displays a special area of memory is dedicated to graphics only. This meniory area is called frame buffer.. It holds the set of intensity values for e screen points. The stored intensity values are retrieved from frame buffer and dtsplayed On the screen one row (scan line) at a time. Each screen point is referred to as a pixel or pel (shortened forms of picture element). Each pixe] on the screen can be specified by its row and column number. Thus by specifying row and column number we can specify the pixel position on the screen. Intensity range for pixel positions depends on the capability of the raster system. It can be a simple black and white system or colour system. In a simple black and whiteComputer Graphics 1-11 _ Introduction to Computer Graphics system, each pixel position is either on or off, so only one bit per pixel is needed to control the intensity of the pixel positions. Additional bits are required when colour and intensity variations can be displayed. Upto 24 bits per pixel are included in high quality display systems, which can require several megabytes of storage space for the frame buffer. On a black and white system with one bit per pixel, the frame buffer is commonly called a bitmap. For systems with multiple bits per pixel, the frame buffer is often referred to as a pixmap. [ Vector Scan Display Raster Scan Display ‘Jn ‘vector scan display the beam is moved | 1. In raster scan display the beam is moved between the end points of the graphics all_over the screen one scan line ata primitives — time, from top to bottom and then back to f - 2. Vector display flickers when the number of | 2.In raster display, the refresh process is Primitives in the buffer becomes too large. independent of the complexity of the a we a WOR Se 3. Sean conversion i not required. 3.Graphics primitives are specified in terms erry * of their endpoints and must be scan converted into their corresponding pixels in the frame buffer. 4. Scan conversion hardware is not required. 4.Because each primitive must be Tak ‘scan-converted, real time dynamics is far more computational and requires separate scan conversion hardware. 5. Vector display draws a continuous and | 5. Raster display can display mathematically smooth lines. smooth lines, polygons, and boundaries of curved primitives only by approximating them with pixels on the raster grid. 6. Cost is more. 6. Cost is low. 7.Vector display only draws lines and | 7. Raster display has ability to display areas characters. On a _|__ filed with solid colours or patterns Table 1.1 1.4.3 Simple Raster Display System The Fig. 1.10 shows the simplest and the most common raster display system organisation. It consists of CPU, system memory, frame buffer, video controller and monitor. The CPU and system memory work exactly in similar manner as in nongraphics computer system. However, in raster scan displays a special area of memory is dedicated to graphics only. This memory area is called frame buffer. It holds the set of intensity values for all the screen points. The video controller retrievesComputer Graphics 4-12 _ Introduction to Computer Graphics the stored intensity values from frame buffer and displays them on the screen one row (ecan line) at a time, typically 50 times per second. Peripheral devices Fig. 1.10 A simple raster display system architecture Many times, the simple raster display system does not provide separate frame buffer. In that case it is stored anywhere in the system memory and the video controller accesses the system memory via the system bus. This is illustrated in Fig. 1.11. Fig. 1.11 Simple raster display system architecture without separate frame buffer We have seen that the video controller receives the intensity information of each pixel from frame buffer and displays them on screen. Let us see the internal organisation of a video controller The Fig. 1.12 shows the internal organisation of a video controller. It consists of raster-scan generator, x and y address registers, and pixel value register. The raster-scan generator produces deflection signals that generate the raster scan. TheComputer Graphics 1-13 Introduction to Computer Graphics raster scan generator also controls the x and y address registers, which in turn define the memory location to be accessed next. The frame buffer is addressed in x from 0 to Xma and in y from Yin. (top scan line) to 0. During each fetch the pixel value is read and is used to control the intensity of the CRT beam. eee Horizontal and vertical deflection signals Bete deze Intensity or colour Fig. 1.12 Simple organisation of the video controller At the beginning of the scan, X address register is set to zero and the Y address register is set to zero and the Y address register is set to Ym During scanning the first scan line X address is incremented from 0 to Xpq in steps of one. At each X address, pixel value is fetched and is used to control the intensity of the electron beam. After the first scan line, the X address is reset to zero and the Y address is decremented by one. The process continues until the last scan line (Y= 0) is generated. In case of colour display memory gives RGB colour values for the pixel. Let us consider the display of resolution of 640 pixels by 480 pixels and refresh rate is 60, ie. each pixel is displayed 60 times in a second. Then the time available for displaying a single pixel can be given as 1 T= ‘GOR 480 XE = 54 nanoseconds Now, if we consider one memory access to the frame buffer for each pixel to be displayed then we have to access pixel value within 54 nanoseconds and this is not Possible because typical RAM memory chips require access time of 200 nanoseconds. To solve this problem video controller fetches multip!e pixel values in one memory cycle. If it can access 16-bits, 16-bits are loaded into a register on the video controller, then are shifted out to control the electron beam intensity, one each 54 nanoseconds. Therefore, memory cycle time is elongated upto 16 pixels x 54 ns/pixel = 864 nanoseconds. This time is quite enough to access pixel data from memory. InComputer Graphics 4-14 Introduction to Computer Graphics 864 nanoseconds, 200 nanoseconds are used for accessing frame buffer and remaining time i.e. 664 nanseconds are available for CPU to access memory. So far we have assumed that 1-bit per pixel bitmaps. But in practice, multiple bits per pixel are used to get additional control over the intensity of each pixel. For example, 2-bits give four intensity levels, 3-bit gives 8 intensity levels and so on. In colour displays, 24-bits per pixel are commonly used, where 8-bits represent 256 levels for each colour. Here, it is necessary to read 24-bits for each pixel from frame buffer. This is very time consuming. To avoid this video controller uses look-up-table (LUT) to store many entries of pixel values in RGB format. With this facility, now it is necessary to only read index to the look-up-table from the frame buffer for each pixel. This index specifies the one of the entries in the look-up-table. The specified entry in the look-up-table is then used to control the intensity or colour of the CRT. Usually, look-up-table has 256 entries. Therefore, the index to the look-up-table has 8-bits, and hence for each pixel frame buffer has to store 8-bits per pixel instead of 24-bits. Fig. 1.13 shows the organisation of a video look-up-table. of ifofofofofo] 1] 4147-65 255 0 Red Green Blue Fig. 1.13 Organisation of a video look-up-table Advantages of simple raster display system : 1. It is simple. 2. It is economical.Computer Graphics 1-15 __ Introduction to Computer Grap! ‘sadvantages of simple raster display system 1. Scan conversion is done using software. Software-based slow, because software has to calculate the address of ea must be translated into a memory bit-within-byte pair. scan conversion is ich pixel and then it address consisting of a byte and 2. Software-based scan conversion also slows down the overall speed of user interaction with the application. 3. In this architecture, as the addressability or the refresh rate of the display increase: the number of memory accesses made by the video controller also increases, thus decreasing the number of memory cycles available to the CPU. This results slow down of CPU, 1.4.4 Raster Display System with Peripheral Display Processor In some graphics systems a separate processor is used to interpret the commands in the display file. Such a processor is known as display process: ‘or. Display processor access the display information, Processes it once during every refresh cycle. In the raster scan display systems, the Purpose of display processor is to free the CPU from the graphics routine task. Here, display processor is provided with separate Memory area, asishown in..the, Fig.114. The main. task of display processor is to digitize a picture definition given in an application program into a set of pixel-intensity values for storage in the frame buffer. This digitization process is known as scan conversion. Display processor memory Video controller Display Processor System memory VO devices Fig. 1.14 Raster scan system with a display processorComputer Graphics 4-16 Introduction to Computer Graphics 1.4.5 Additional Display-Processor Functionality Display processor are also designed to perform a number of additional operations. These operations include © Generating various line styles (dashed, dotted, or solid) ¢ Display colour areas «Performing certain transformations and * Manipulations on displayed objects By adding the more functionality to the display processor, CPU can be offloaded. Ultimately, the display processor becomes another general-purpose CPU doing general interactive graphics work. 4.4.6 Raster Display System with Integrated Display Processor The architecture of the raster display system with integrated processor is as shown in the Fig. 1.15. It consists of display processor, CPU, system memory and the video controller. All these are on the system bus and all access the system memory. Thus, such systems are known as single-address-space (SAS) raster display systems. In such architectures, the size of the frame buffer is held in registers, making double-buffering a simple matter of reloading the register. Here, the results of scan conversion can go either into the frame buffer for immediate display, or they can be stored in the system memory. Peripheral devices Fig. 1.15 Single-address-space (SAS) raster display system architecture with integral display processor Disadvantages of SAS architecture 1. Since all memory accesses are to the system memory, there is a serious problem of bus contention.Computer Graphics 1-17 __ Introduction to Computer Graphics 2. This problem can be partially solved by dedicating a special portion of system memory as a frame buffer and providing a second access port to the frame buffer from the video controller as shown in the Fig. 116. Another solution to contention problem is to use a CPU having instruction cache or data cache memories, thus reducing the CPU’s dependence on frequent and rapid access to the system memory. Fig. 1.16 SAS architecture with separate access to frame buffer 3. Most of the commonly used CPUs such as Motorola 680 x 0 and Intel 80 x 86 families and various RISC processors have a virtual address space. In such cases, memory addresses generated by the display processor must go through the same dynamic address translation as other memory addresses. This is time consuming. Therefore, display instructions which access the frame buffer must distinguish between the Kernel and application address spaces and operating systems device driver should access frame buffer directly. 1.4.7 The Video Controller The most important task of the video controller is to refresh the display continuously. There are two basic types of refreshing display : interlaced and noninterlaced. The interlaced scanning is used in broadcast TV and in raster displays designed to drive regular televisions. In this scanning, refresh cycle is divided into two fields : all odd-numbered scan lines are displayed in the first field, and all even-numbered scan lines are displayed in the second field. The main purpose of interlaced scan is to avoid flicker (It is the effect of discontinuity between the two picture frames). The Fig. 1.17 shows the concept of interlaced scanning. Starting from first line, the alternate lines are scanned in the first part (or field) of the frame and remaining lines are scanned in the second part (or field) of the frame. Thus one picture frame is divided into two picture fields. Thus in the 625-line system, one picture frame contains