Development of Displays: Getting to See 3D

Abstract

Display technology has had three basic phases in its history starting with monochrome vector CRT, going through monochrome raster CRT to color raster CRT, and then using color raster LCD and OLED. CRTs are either vector, or raster. A vector, or electrostatic CRT, (sometimes referred to as “random scan”) was first used as a used in computer systems from the 1940s to as late as the late 1980s and were replaced by electromagnetic deflection raster scan displays (sometimes referred to as “all points addressable” APA display). Vector displays were round, and usually large, 20–24-in. in diameter. Raster displays, based on TV tubes, were rectangular, ranging in size from as small as 9-in., up to 27-in., in the 2000s 30-in. raster-scan LCD computer monitors became available. Generally, if you see a picture of an old computer you can tell what kind of display it has by its shape. A special class of vector display was developed in 1968 called a storage tube. The Direct View Storage Tube emerged as a graphics screen that maintained an image without requiring refreshing (However, the entire screen had to be redrawn for any change). Vector graphics terminals, which evolved from oscilloscopes, required constant refreshing of the image—one of the reasons the storage tube display terminals were so popular.

The first production color picture tube was a 15-in. round screen CRT, made in 1954 by RCA. Raster scan displays for graphics were initially limited to X–Y resolutions such as 256 × 256 to 512 × 512, or 480 × 640 due to the cost of the memory in the frame buffer, and they were slow due to the difficulty of the processing required for scan conversion.

Display standards for the PC were developed by IBM and can still be found in use today (in the form of the venerable VGA standard). Today its alphabet-soup with standards such as DVI, HDMI, DP, and others.

Projectors have gotten small and low cost making very large displays using multiple projectors possible. New technologies in the form of nano-crystals called quantum dots will probably be the final surface display technology until holographic displays become practical.

Appendices

Appendix

Here are listed items that may be useful in understanding more about the industry and technology.

7.1.1 Pioneering Companies

Table 7.1 Pioneering hardware companies in computer graphics

In the late 1960s and early 1970s, a number of new computer graphics companies were organized. Listed above are just a few of these companies and the dates on which they were founded or on which they became active in computer graphics (Table 7.1).

7.1.2 Calculation of Monitor PPI

Theoretically, PPI can be calculated from knowing the diagonal size of the screen in inches and the resolution in pixels (width and height). This can be done in two steps:

• 1. Calculate diagonal resolution in pixels using the Pythagorean theorem:

  $$ {d_p}=\sqrt{{w_p^2+h_p^2}} $$

• 2. Calculate PPI:

  $$ \textit{PPI}=\frac{{{d_p}}}{{{d_i}}} $$

  where

  • d _p is diagonal resolution in pixels,

  • w _p is width resolution in pixels,

  • h _p is height resolution in pixels and

  • d _i is diagonal size in inches. (This is the number advertised as the size of the display).

7.1.3 Moore’s Law

“The number of transistors and resistors on a chip doubles every 18 months”. By Intel co-founder Gordon Moore regarding the pace of semiconductor technology. He made this famous comment in 1965 when there were approximately 60 devices on a chip. Proving Moore’s law to be rather accurate, four decades later, Intel placed 1.7 billion transistors on its Itanium chip.

In 1975, Moore extended the 18 months to 24 months. More recently, he said that the cost of a semiconductor manufacturing plant doubles with each generation of microprocessor.

Appendix

With the introduction of the IBM PC, the industry got its first display interface standard; the Monochrome Display Adapter (MDA). The Color Graphics Adapter (CGA) in 1981, a limited color graphics controller that drove a monitor with fixed intensity steps (referred to as “digital”), quickly followed it. Those standards were used up until the late 1980s, with independent companies like Artist graphics and Hercules going around them.

7.2.1 EGA

The Enhanced Graphics Adapter (EGA) developed by IBM in 1984, was an innovation. Not only did it support the previous CGA modes it offered 640 × 350 resolution with the ability to display 16 colors simultaneously. The displayable 16 colors were chosen (by the application) from a palette of 64 colors (which is 6-bit RGB). The EGA output was for what IBM then called an “analog RGB” type monitor.

The word analog means that the RGB signals can have more than the basic binary levels 0 and 1 (which IBM called the “digital RGB” CGA monitor type). The number of voltage levels or steps defines the number of intensities available for each primary color, and the “primary” colors (Red, Blue, and Green – RGB). So if there were six possible voltage steps for each primary, then there could be 64 combinations of intensity and primaries, and that is referred to as 64 colors.

Despite being called “analog”, colors are produced digitally, so there are binary (quantized) values for every primary RGB signal. Those digital values are converted to analog voltages with a special output amplifier known as a digital to analog converter, or DAC. The DAC’s analog output drives the monitor.

7.2.2 VGA and the PS/2

In 1987 IBM introduced a new line of PCs called the PS/2. Part of the new machine was an analog RGB display adaptor known as the Video Graphics Adaptor – VGA. It had a resolution of 640 × 480 and could support 256 colors simultaneously (8-bit RGB). The VGA adaptor was immediately the new standard, and still lives with us today in the form of its 15-pin connector (Fig. 7.70).

Fig. 7.70

IBM’s 1987 ubiquitous VGA connector can still be found on PCs and monitors

Now the PC industry had three standards, all from IBM, the CGA, EGA, and VGA. What would be next?

In retrospect, it was remarkable, and wonderful that IBM had such influence in the industry. Almost every other company had more advanced technology, and could deliver it quicker. But IBM represented stability, much as Microsoft does today. Therefore, while all the other companies would push the technology, it was left to IBM to set the standards. Also, back in the 1980s the enterprise was the big customer for computers, mainframes and PCs. PCs were coming down in price and slowly being taken home, but the volume buyers was the enterprise customer. And those companies wanted stability and long-term dependability. No other company in the PC market could offer the scale that IBM could.

7.2.3 IBM 8514

History repeated itself and at the same time IBM was introducing the VGA adaptor for the mainstream, it also brought out, on the PS/2, a higher performance, higher resolution adapter called the 8514. The 8514 pushed display technology to 1024 × 768 resolution, with just 256 colors. It was a disappointing product, and years behind what the industry was offering. However, it did establish the resolution standard, still used today, of 1024 × 768, and known as extended VGA – EVGA, and then shortly XGA which is where it is today.

Warninggeeky alphabet-soup stuff to follow – my apologies

7.2.4 VESA

Adaptations were developed to the VGA standard and in 1989 an industry consortium named the Video Electronics Standards Association – VESA was set up to try and wrest away some of IBM’s influence. Two years after the introduction of the 640 × 480 VGA and the 1024 × 768, 8514 adaptor, VESA came out with the Super VGA (SVGA) specification of 800 × 600 resolution. However, it was behind on color range and only specified 4-bit pixels. Each pixel could therefore be any of 16 different colors.

Prior to the ubiquitous VGA connector, the display industry used three BNC (Bayonet Neill–Concelman) RF connectors (also known as Co-axial connectors), one for each primary color Red, Blue, and Green, for multi-color signaling from the graphics controller to monitors and projectors (Fig. 7.71).

Fig. 7.71

BNC (co-axial) connector (Courtesy of Meggar (CC BY-SA 3.0))

Sometimes (in the early designs) there was a fourth connector for synch. In later designs, the synch signal was superimposed on the green signal. The RGB BNC connections were a holdover from RADAR and TV broadcasting and many of the standards developed for TV and can still be found in some modern display technology (Fig. 7.72).

Fig. 7.72

BNC video connectors (© 2012 The Render Q)

VGA, SVGA, and XGA (as the IBM 1024 × 768 specification came to be called) all use the VGA connector and are analog signals. The analog signaling resolution was pushed up to 1600 × 1220 (called UXGA) and some brave developers even got it to 2048 × 1536 – beyond HDTV. Nevertheless, the durable and venerable VGA interface was analog and suffered the traditional stability problems of analog (e.g., shifting of voltage levels (i.e., color intensity) over time and/or due to heat, and electrical noise interference). Also there was no way to get any information about the monitor itself.

7.2.5 DDC/SDIC

When VESA took over the specification management of the VGA, they added a display data channel (DDC). Several versions evolved but the main idea was that the display controller should be able to know something about the characteristics of the monitor and exert some control over it. In 1996, VESA extended the DDC to become the EDID – Extended Display Identification Data channel. However, although it was an industry standard not all monitor manufactures deployed it, or followed its specifications to the letter, and there was no certification process so it was optional, and therefore not very useful.

7.2.6 DVI

The Digital Visual Interface – DVI (many people think the “V” in DVI stands for video) is based on the PanelLink technology developed by Silicon Image, Inc. in 1995. The company gave the technology given to the industry, which formed the industry consortium, the Digital Display Working Group (DDWG) to replace the “legacy analog technology” VGA connector standard.

Silicon Image originally develop the PanelLink technology to provide an interface between set-top boxes (STBs), DVD players and TVs. The design was quickly adapted to PCs and other computers and the need developed to move it from a proprietary design to an industry standard.

In 1998, the DDWG consortium released its first industry specification. DVI solved many problems, offered extendibility into very high frequencies (and therefore resolutions), and moved the industry one-step closer to an all-digital system.

DVI uses uncompressed video only, and High-bandwidth Digital Content Protection (HDCP). Encryption is optional. To get the very high resolutions such as 30-in. displays with 2560 × 1600 resolutions, a dual link solution is required.

DVI also brought a new connector design, which has pins in it for older analog VGA signals, single and dual link transmitters, and monitor signaling and control. With all those combinations, the DVI connector on a device can be one of four types, depending on which signals it implements:

• DVI-D (digital only)

• DVI-A (analog only)

• DVI-I (integrated, digital & analog)

• M1-DA (integrated, digital, analog & USB) (Fig. 7.73)

  Fig. 7.73

  

  DVI connector types (Copyright free, image released into the public domain by Hungry Charlie)

However, as good as DVI is, it too lacked a certification and regulation body and so the quality of the signal, known as the TMDS “eye” could vary widely from supplier to supplier. In addition, there is no provision for audio in the DVI specification although some people argue it could easily be superimposed on one of the signal lines.

7.2.7 HDMI

Building on their success with DVI, Silicon Image was a founder and key developer of the High Definition Multimedia Interface (HDMI) and a final specification was introduced in December 2002. The major consumer electronics manufacturers such as Hitachi, Philips, Sony, and Toshiba ratified it. Based on TDMS like DVI, HDMI is backwards compatible with DVI.

The original HDMI specification (1.0) provided a audio/video connection with a maximum bitrate of 4.9 Gbps, or up to 165 Megapixels/second of video (1080p @ 60 Hz or UXGA) and 8-channel/192 kHz/24-bit audio. DVD content protection was a big issue then and so the first revision (HDMI 1.1) added content protection HDCP (High-bandwidth Digital Copy Protection).

Various other revisions were added for audio standards that were coming out (e.g., DVD audio, super CD audio, Dolby and DTS_HD). The bandwidth was extended to 10.2 Gbps in 2006, and the color space was expanded over the years up to 48 bit RGB and YCbCr. Blu-ray formats were added as well as Blu-ray stereo vision (“3D”) all the while maintain backward compatibility.

HDMI began appearing on graphics add-in boards in 2006, the first one appearing on an ATI board. As PCs became a part of home entertainment system HDMI soon became a standard feature so board were appearing with a VGA, DVI AND an HDMI connector – let no display standard go un-serviced.

7.2.8 DisplayPort

In the fall of 2005, VESA the self-appointed video electronics standards organization formally announced the royalty-free DisplayPort specifications. The arguments in favor of the proposed standard seemed flimsy at best. VESA postulated that because DVI, used in computers, and HDMI, used in CE devices, have a similar physical layer (PHY), it “leads the consumer to believe that such products will interoperate”. However, the original DisplayPort specification, developed by Dell, seemed to be an attempt to get around Silicon Image’s licensing fees for DVI and HDMI. At the time, Dell’s PC group was not an HDMI user but their TV group was. Almost all new mid- to high-end TVs then had HDMI connectors on them. So if Dell thought they would save royalty fees, what about the additional connector and duplicated content protection licensing fees? It was another connector and not one asked for (Fig. 7.74).

Fig. 7.74

DisplayPort (full-size) connector (Courtesy of Oliver Abisys (CC BY-SA 3.0))

Aimed to replace internal LVDS (see above and appendix) links in notebook panels with a unified link interface, DisplayPort incorporated a main link, a high-bandwidth, low-latency, unidirectional connection offering isochronous stream transport – one uncompressed video stream and associated audio. The developers said it was extensible, enabling support of multiple video and/or audio streams. There was also an Auxiliary Channel to provide device control based on VESA’s EDID (Extended Display Identification Data channel, see above) and MCCS (Monitor Control Command) standards.

The Main Link bandwidth of up to 10.8 Gbits/s, equivalent to a data transfer rate of 1080 MBytes/s, uses four lanes; the auxiliary channel features minimal delay, with maximum transaction periods less than 500 μs. Data is transmitted across the DisplayPort interface using a micro- packetized format. This represented state-of-the-art signaling technology and held the promise of being extensible—at last a potentially long-life display standard.

However, DisplayPort was an unwelcomed specification (that would ultimately be saved by Intel’s UDI). In the meantime, there were the lawyers. DisplayPort is the trademarked name of the portable display company (http://www.dis-playport.ca/). In 2007 a private agreement was reached.

7.2.8.1 UDI

Late 2005 saw a new display interface introduced with the premise that the DVI connector was too large to conveniently fit in a laptop, especially a thin-and-light notebooks, and that the venerable VGA connector (which isn’t much smaller than a DVI) has outlived its usefulness in this age of digital everything. The solution was the Universal Device Interface (UDI). A specification developed by Intel quickly led to a SIG (Special Interest Group – the first stage of getting a standard established). Apple joined Intel and the leading graphics add-in board (AIB) companies, ATI and Nvidia, plus several PC suppliers, and even Silicon Image. UDI was compatible with HDMI and supported HDCP copy protection, which always is a big deal.

UDI provided higher bandwidth than its predecessors (up to 16 Gbit/s in its first version, compared to 4.9 Gbit/s for HDMI 1.0).

7.2.8.2 The Merger

UDI had qualities DisplayPort lacked such as a micro-packet protocol, which would, allow an easy expansion of the standard. It could support multiple video streams over single physical connection, and it could handle long-distance transmission over fiber optic media. In addition, UDI would support internal chip-to-chip communication. It supported RGB and YCbCr encoding formats but so did DisplayPort. However, Intel knew it wouldn’t do the industry any good to fragment it further and in 2007 led the merging of DisplayPort 1.0 and UDI. In November 2006, VESA announced DisplayPort 1.1 (later approved in ’07), with Intel on-board as a supporter and task group member. There was peace in the valley at last.

With the new DisplayPort, PC and graphics AIB suppliers began offering DisplayPort on their products, and often with a dongle that could convert from DisplayPort to DVI or HDMI. The industry was slowly evolving toward a single connector.

7.2.8.3 Mini DisplayPort

The next step was make a mini-me – Mini DisplayPort is a miniaturized version of the DisplayPort interface. Apple first publicly announced the mini-DP in October 2008 for its new thin notebooks. However, unlike its Mini-DVI and Micro-DVI predecessors, Mini DisplayPort was capable of driving resolutions up to 2560 × 1600, commonly used with 30-in. displays. With a suitable adapter, Mini DisplayPort could drive displays with a VGA, DVI, or HDMI interface (Fig. 7.75).

Fig. 7.75

Mini DisplayPort connector is about one-fourth the size of a standard DisplayPort connector

Apple licensed the Mini DisplayPort connector with no fee to VESA and in January 2009, VESA announced that Mini DisplayPort would be included in the DisplayPort 1.2 specification (Fig. 7.76).

Fig. 7.76

Six displays driven by one AIB made possible by mini DisplayPort (©2012 Advanced Micro Devices, Inc)

Perhaps one of the most impressive use of the mini DisplayPort connector was that done by AMD with a graphics AIB that drove six displays at high-resolution simultaneously (AMD called this their Eyefinity technology).

7.2.9 USB

However, the truly universal I/O connector found on all PCs, most mobile phones, game consoles, and TVs is the Universal Bus Standard connector – USB.

USB is a lower bandwidth (than video) serial signaling technique that has been used to connect almost everything to a PC including mice, keyboards, cameras, charging of mobile device like phones and cameras, and external disk drives.

The USB 1.0 specification introduced in 1996 and had a data transfer rate of 12 Mbit/s. In April 2000, the USB 2.0 specification extended it to 480 Mbit/s (Fig. 7.77).

Fig. 7.77

The Universal Serial Bus connector (Courtesy of Afrank99 (CC BY-SA 2.0))

However, even before USB was extended, companies began experimenting with compressing the image in the graphic’s frame buffer, sending it out over the USB, and then decompressing it at the display. DisplayLink Inc. (formally named Newnham Research), founded in 2003 in Cambridge UK, introduced their first commercial product in 2006 the DL-120 and then the DL-160 USB 2.0 graphics device, in January 2007. Now bulky VGA or DVI cables weren’t needed, and with USB powered hubs, the display could be a great distance away from the computer at a minimum cost (Fig. 7.78).

Fig. 7.78

DisplayLink USB to DVI dongle (©2012 DisplayLink)

However, the compression wouldn’t support actual streaming video – that had to wait for the introduction of USB 3.0 in 2010. Several peripheral suppliers like Logitech, Evga, Kingston, etc. and some PC suppliers like Dell, adopted the technology. Dell employed it wirelessly so a second or remote monitor could be attached to a laptop without any cables (other than power).

Since most monitors were being built with USB hubs it seemed as if the monitor suppliers might finally be able to reduce cost and complexity and offer a display with just one connector; a few did (Fig. 7.79).

Fig. 7.79

Acer’s B223 Vista monitor with just a USB connector (© Acer)

Acer’s B-series Vista Aero-compatible USB computer monitor has DisplayLink technology integrated, and features 22-in. widescreen panel with 1680 × 1050 resolution, and 32-bit true-color graphics. In 2012 the unit sold for $438.

7.2.10 The Connectors

However, ever afraid of disappointing a single potential user rather than reduce the number of connectors on a monitor, they got expanded, and in 2010 you could buy a monitor that had VGA, DVI, DisplayPort, HDMI, USB, A/V, and even S-Video connectors – and most projectors and large screen TVs were similarly equipped—let no signal go unwelcomed (Fig. 7.80).

Fig. 7.80

Rear panel of high-end projector (Courtesy of Vivitek)

You could even find systems with RGB BNC connectors.

7.2.11 Those GAs

Starting with the CGA from IBM, the attachment of a prefix to designate its resolution range came into being when monitor and TV manufactures wanted to differentiate their screens and thought numbers like 1024 × 768 would be too confusing for the (they thought) dumb consumers, and meaningless and confusing acronyms would be easier. So the industry was introduced to WXGA, SXGA, and WSXGA as if that would mean something.

There are three parameters for the physical characteristics (not physical size) of a display, its horizontal and vertical resolution, and its aspect ratio. A partial list is offered here. It’s partial because the screen manufactures are coming up with new versions very often (Table 7.2).

Table 7.2 Popular display resolutions and their names

Fig. 7.81

DVI TMDS “eye” (© 2004 Silicon Image, Inc)

While the mainstream PC graphics were being developed and fought over, higher-performance graphics for game playing, professional graphics for CAD and visualization, and medical and scientific instrumentation were also being developed.

Computer Graphics Course – http://www.gomezconsultants.com/CSE5280/GraphicsHardware.html

TMDS – Transition Minimized Differential Signaling is one DVI link that consists of four twisted pairs of wires and is used to transmit 24 bits per pixel. The timing of the signal almost exactly matches that of an analog video signal. The term “Differential is the magic and it balanced the line so that almost no electrical noise can get into it.

TMDS is similar to Low-Voltage Differential Signaling (LVDS) in that it uses differential signaling to reduce electromagnetic interference (EMI) which allows faster signal transfers with increased accuracy. (LVDS is what’s used in the IEEE 1394 interconnects).

As the signal switches from one state to another it ramps up and when the signal is view on an oscilloscope it looks like an eye.

As of 2010 DVI was the primary video signaling standard and found on all monitors and new TV sets (Fig. 7.81).

7.2.12 Literature

Before Newman & Sproull wrote their classic CG book in 1973, there was a generation of display graphics books that preceded it. They carried titles such as:

• William A. Fetter Computer Graphics in Communication (1964)

• Harry Poole s Fundamentals of Display Systems (1966)

• Fred Gruenbergers Computer Graphics Utility/Production/Art (1967)

• Murray Milnes Computer Graphics in Architecture and Design (1969)

• Parslow, Prowse, and Greens Computer Graphics Techniques and Applications (1969)

• Hortons Data Display Systems (1970)

• Sol Sherrs Fundamentals of Display System Design (1970)

• David Princes Interactive Computer Graphics for Computer-Aided Design (1971)

• Jasia Reichardts Cybernetics, Art and Ideas (1971).