In introducing Dr Tonkin, Mr Neil Clarke of the Computer Centre was careful to differentiate between the essentially desktop video systems (for one-to-one communication) which Bruce would be focussing upon, and the larger scale lecturing systems employing video projectors which would receive attention at the April HEPCIT presentation. He drew attention to the fact that the major communications companies and the major computer vendors are all pouring billions of dollars into development of digital video compression, storage and transmission systems, because of the obvious monetary returns that would follow.
Bruce began by explaining where ANSPAG fits into the Centre for Telecommunications and Information Engineering and he pointed out that he was essentially representing that Centre in his presentation because he would be talking about technologies developed across the Centre and not just ANSPAG.
Moving into the topic, Bruce explained that digital video is simply another component of the mix of media which can be manipulated and processed by computers, in contrast with the analogue form of video which is broadcast into our homes and recorded on standard VCRs. In this context, digital video is a component of `multimedia', and a key factor in allowing this integration of video into the PC environment is the ability to use video compression. This technology reduces data storage requirements, and the cost of transmission of the reduced quantity of data. This is very important when considering remote access, for example, inter-campus linkages or access by remote students at home in Australia or even overseas. The advent of video compression has enabled what Bruce called `personal video services', that is, video delivered to you personally, rather than to a thousand or million other recipients. This means that, at one end of the scale, collections of films could be stored in computer systems for network access worldwide by individuals, or at the other extreme, video conferences could be held between two PC users.
Bruce noted that until recently, video conferencing has involved the use of very expensive systems installed in special purpose conference rooms, using proprietary equipment and technique. In general, participants have had to pre-book these facilities and travel to a conference centre to take part, and for an inter-city link it could cost up to $600 per hour. An advantage of these systems is that attention is paid to providing good audio between the centres. Standards are now being developed for these systems so that equipment products from a range of vendors can inter-operate successfully. The end result of this standardisation is that the cost of systems is reducing through competition, in the same way that we have seen in the fax machine market. Many individuals now can afford to have a fax machine, and the same trend will be seen in personal video.
Bruce next moved on to outline the typical configuration of a desktop video system, with the aid of projected graphics. A design consideration is whether the compression / decompression should be handled by hardware, that is, a separate `codec' card in the PC, or whether this can be handled by the PC's CPU via dedicated software. At this stage this decision raises an issue of cost versus quality, special expensive hardware being needed at present to achieve acceptable codec rates. But as the CPU power of PCs increases, it is expected that software will be used to handle this aspect at much lower cost.
Other issues to be addressed include
- whether the communication be via the telephone network or a computer data network,
- whether to use a standards based system, a proprietary system, or even a public-domain system as being used at present on the Internet.
At present, proprietary systems are very much aimed at computer data networks, there being no current standards in that area. The only standards based systems are across the digital telephone networks, such as the ISDN network employed by Telecom.
At this point, Bruce demonstrated a typical desktop video system, based on a standard PC containing a special codec hardware board to which was connected a video camera and which communicated via a dedicated self-contained digital telephone system to a similarly configured system at the rear of the lecture theatre. The limitations of the video compression and transmission systems were evident here, in those parts of the picture that were changing. But considering that the link was via a telephone system, the resulting image was acceptably good. However, this is tempered by the cost - the add-on hardware that Bruce demonstrated costs $8000(!), but these prices are dropping rapidly. An advantage of this system is that it is standards based, and is completely compatible with the University's main room-based video conferencing system, and most other systems available today.
As mentioned above, software will be used in the future to handle the video compression, and the Centre for Telecommunications is working towards developing software to transmit video across a local area network and then a connection between the LAN and the existing telephone-based network. But at this stage there are no standards for the analogue telephone network, or for LANs, only the digital telephone network. In the future we can expect to see PCs already equipped to communicate video to and from the telephone network and to LANs. The new Apple Macintoshes, for example, have a video analogue to digital converter built in, as well as provision for sound.
The big advantage of using PCs for video communication is that you have access to applications already running on the PCs, such as word processors and presentation tools, and the output from these applications can be shared between the interconnected PCs. The main disadvantage at present is in the handling of audio. Audio break-up is a common problem, and strange echo effects can be introduced if the output from a speaker is picked up by a neighbouring microphone.
Apart from video on computer and telephone networks, the other main area of activity is in use of the Internet. Some attention is being given worldwide to the video broadcasting of events such as seminars via the Internet. A notable example of this technique is that NASA has a cable TV channel in the USA which carries 24 hour broadcasts of space shuttle missions. This is now also being sent to the Internet in digital format. This works fine in one-way transmissions, but audio problems occur in two-way links if the transmission is being handled by satellites where a two second delay occurs. In most Internet usage, the software employed is public domain, and the codec and transmission rates limit the video frame rate to about 3-4 frames per second, with poor motion representation. In about five years time, with much more powerful CPUs available, this problem should largely disappear.
At this stage, Bruce summed up the options available at present for video conferencing, and recommended that at least for the next year, hardware based systems are needed for acceptable quality in serious applications such as teaching. He then outlined some of the applications that could employ desktop video to good effect, and Bruce's summary of these together with his elaboration of many of them demonstrated his obvious desire to see the technology achieve its potential.
He concluded his presentation with, firstly, a hardware based demonstration of how two PC users, via video technology across a digital telephone network, can communicate images to aid in discussion of a subject of common interest, and also to share and discuss applications between the PCs. In this, he was assisted by Kevin Dillon from the Network Systems Group and Russell Lang from the Video Communications Group. They began by discussing the possible modification of a network circuit card which Russell showed to the camera. Some audio problems were encountered here, owing to sound leakage from the booth at the rear of the theatre, but the video was acceptably good. They then went on to share an application in Microsoft Powerpoint, and this demonstrated clearly that high quality video images are achieved when data of this kind is transmitted from one PC to the other. Unfortunately, Kevin's hand waving technique, clearly obvious to the audience, was not successfully communicated to the other end!
Secondly, they then showed a similar connection using a software codec technique, in which a `whiteboard' was shared between the users. In this, highlighting of parts of the image by boxing and circling some of the existing graphical elements was successfully transmitted to the other end. Some difficulty was encountered with the lighting here, and Bruce suggested that those interested in this technique should see it working properly in his laboratory.
This ended the presentation, which was very successful in outlining what can be achieved now in desktop video, and where the technology is heading. It also provided a useful pre-cursor to the April HEPCIT, in which the technologies to be used in the Berwick Project will be the focus of attention.
The HEPCIT Steering Committee extends its thanks to Dr Tonkin and his helpers for the considerable effort they expended in making this first HEPCIT presentation for 1995 the success that it was.