Search
Another full issue! Read about the meeting for Graphics Supporters, and the CHEST deal for the visualization software AVS. There are reports on the Rendering Workshop and SIGGRAPH92. Graphics Around the Country looks at Graphics in SERC. Thanks to all those who have provided information and submitted reports. We welcome news, reports. and articles for future issues.
As advertised in the last Graphics Newsletter, we are holding a gathering for graphics support staff. This is being held at Loughborough on 14 and 15 December. The programme is being put together by Alex Nolan of the University of Edinburgh. Topics so far on the provisional programme are: how packages fit together; selecting packages and graphics programming tools; colour; visualization; X and related issues; coping with the complexity of supporting graphics.
Now is the time to make sure you have the date in your diary. The event is being funded by the Information Systems Committee of the UFC as part of the AGOCG budget so you only have to find the cost of your travel. Places are limited though, so if you want to make sure of an early copy of an application form please E-mail a postal address to me (I seem to still believe in paper!)
What role should AGOCG play in the area of multi-media? That is the question which will be addressed at a workshop on 25 November. There are clearly lots of interested groups of people in this area and we need to establish our role in relation to others to try and gain a common direction in the HE community. If you feel you have an input to make then please E-mail me and I win ensure you get the details.
This is being held in January and will be multi-disciplinary. See inside this newsletter for the details. It should be very informative and productive for students. The students will only have to pay the cost of their accommodation and notes as the fixed costs are being met for them. This is a bargain! Get your students to come.
This is progressing well and we hope to conclude the evaluation in the near future. The details of what has been going on are reported by Steve Morgan in this issue. If you want to ensure an early view of the report of the evaluation please E-mail me giving a postal address.
This exercise has faltered somewhat due mainly to contractual difficulties. CHEST are now talking to one supplier only. The second supplier alluded to in the last Newsletter, withdrew.
Version 3 of Cricket Graph for the Macintosh is to be distributed to sites in late September '92. The new version for Windows has been delayed and is not now expected before December '92. The working party will be reviewing the training materials in the light of the new facilities.
At the last AGOCG meeting (May 8 1992), a recommendation was made to hold a meeting of parties interested in Image Processing/Analysis in the community. This meeting was subsequently held on July 1st, 1992 in London with 14 attendees.
The meeting agreed that software currently available in the Image Processing area can be put in 2 fairly distinct categories:
There was general consensus that now was the right time to evaluate the market and make a recommendation for a CHEST deal.
The meeting first took general comments on the RFI from all attendees. The RFI was felt to be reasonably comprehensive but several shortcomings were noted especially in the areas of vision and medical imaging.
Fred Hopper presented his paper evaluation of the products that had been submitted by suppliers in response to the RFI. From this it was possible to compile a short-list of 6 products which it was thought were worthy of evaluation.
Several public domain products (e.g. Khoros, GRASS) were also discussed and it was agreed that these packages should be positioned relative to the software under consideration.
The meeting also noted the importance of links to other packages, especially UNIRAS and ARC/INFO which are available under existing CHEST deals. It was agreed that evaluations should assess capabilities in this area. More recently developments regarding the new CHEST deal for AVS and its capabilities in the image processing area have been noted.
It was decided to evaluate all short listed products on as many platforms as possible by mid-September. All the major workstation suppliers (SUN, SG, IBM, DEC, HP, ...) were covered by sites represented at the meeting and so it was possible to assign suitable packages/platforms for evaluation covering a fair range of combinations.
A meeting was arranged for September 14/15, 1992, at which the reports would be drawn together and the way forward decided upon. It is hoped to compile a report detailing the results of the evaluation and make this report available as an AGOCG Technical Report.
An E-mail discussion group has been set up (AGOCGip@uk.ac.mailbase) to encourage discussion on the evaluation and IP in general. To join the list send the single line:
subscribe agocg-ip <first_name> <second_name>
to:
mailbase @ uk.ac.mailbase
There is now available for the UK academic community a CHEST deal for the Application Visualisation System (AVS) software which is produced by Advanced Visual Systems Inc.
The deal has reduced the cost of the software by means of a site license which allows a site the option of installing AVS on ALL the following machine ranges:
There will be more information concerning the inclusion of DECstations as part of the CHEST deal around September/October 1992.
Each implementation takes advantage of the graphics sub-system on a particular machine range as well as providing an interface to the X window system which allows it to execute as an X11 client.
The Advisory Group on Computer Graphics (AGOCG) commissioned an evaluation of visualisation systems which started in June 1991. The systems evaluated were AVS 3, apE 2.1 and Khoros 1.0 and consisted of a paper study, a number of case studies and a usability study carried out on each system. The evaluation was completed in January 1992 and the recommendations were to promote the use of visualisation systems via AVS while reviewing the situation in late 1992. More information can be found in the technical report detailing the evaluation which is now available as an AGOCG Technical Report (no 9).
AVS is a system which allows users to visualise their data by constructing applications from a series of software components called modules. Each module performs a specific task and some of these include:
AVS provides a number of generic data types onto which application data can be mapped and imported into the AVS system. These data types provide support for images, arrays of data, geometry and chemistry applications. There is also support for unstructured data where data is associated with discrete objects. This is used to import the results from applications such as finite element analysis.
A key feature of the system is that it is extensible by the user as additional modules can be written, in C or FORTRAN, and integrated into the system.
The terms of the deal will have been sent by CHEST to all their site contacts. If you are interested in the deal at your particular site it is important that you convey this interest to your local CHEST site contact The identity of your local CHEST contact can be found by either contacting your computer centre or by consulting the list of CHEST site contacts on the NISS bulletin board (CALL NISS.BB from a PAD and consult section D3E).
There is also a news group for AVS (comp.graphics.avs), an anonymous FTP site containing public domain modules and a forthcoming introductory course on AVS on the 28 October. For more details and any queries please feel free to contact either Steve Larkin or Julian Gallop.
There will be a one day introductory course to AVS held at the University of Manchester on Wednesday 28 October 1992 presented by Steve Larkin, AGOCG Visualisation Support Officer. It will be aimed at the user with little or no experience of AVS and includes the following topics:
The course will not include more advanced topics such as module writing but a later course is planned for Spring 1993 to cover these.
The course is open to all UK academics and no charge will be made. It will start at 9.30 and finish at about 4.30 with the local arrangements and travel information being forwarded to each attendee nearer the date.
As the course will involve practical sessions using AVS throughout the day the number of attendees will be limited and places allocated on a first come first served basis.
Please include your name, institute address, telephone number and E-mail when registering.
The opening day's conference newspaper told us we were not in Barcelona but in Chicago - though the conference was going to be of Olympic proportions. Indeed it was. This was my first Siggraph and to be honest it was a bit daunting. There must be a critical mass of people beyond which it is very difficult to meet the people you do know, never mind making new contacts. Siggraph is well beyond that limit.
I found the exhibition probably the most worthwhile part of the event There were over 200 vendors. Much of it was animation oriented, software and hardware (video boards etc). It was interesting to talk with AVS about the recent CHEST deal and to see that DECstations would be included shortly. Perhaps more interesting was the news that Silicon Graphics Explorer is to be ported to other platforms, primarily through a tie up with NAG. Not to be left out, the IBM Data Visualise package is also being made available on other platforms. As we expected, this is proving quite a competitive field.
Special exhibit areas showed scientific applications being run in a distributed way between the show and link ups with scientists in the Supercomputer Centres; there was also a live PEX demo round the show floor between vendors. The virtual reality demonstrations had such long queues that there was no way I could be bothered to wait - over an hour and a half to get in. Although I am no expert in the making of computer animations, the film show was fun. There was much appreciation for a clip called Gentle Breeze showing wind blowing some curtains and a tablecloth. The Mandlesplat video was funny, but my favourite had to be the basketball ring running away from the Michael-Jordan-like player coming at it to slam-dunk (technical term!!).
Bob Lucky of AT and T opened the conference and suggested that the networks are out there looking for work and that images and communications have found their joint moment in time. The availability of standard compression techniques was noted as being of importance and this was come back to again and again at various things I attended. We are the providers of the bits that will justify the network provision!
I attended one of the tutorials on networks and visualization. I was disappointed in the balance of this which concentrated more on specific applications rather than helping me uncover the mystique of the networking jargon which the course outline had suggested. I think there is a real need for an understanding of the capabilities that are around today and such a tutorial topic is clearly of value. The first half of the tutorial could have been expanded to the full day for my taste. I guess this is not unusual but I did fill in my form!
I attended some good panel sessions which I found interesting and informative. The one getting all the press coverage - even in the Guardian - was the packed out PEX versus GL battle. My own perspective on this riotous event (amazing for 8.45 am) was that the most convincing arguments came from the PEX side. The speakers from Sun and HP were very much more convincing on the technical side, compared with the knife twisting from Silicon Graphics. The commitment to open solutions for interoperability supported by the PEX consortium is more convincing to me. I have a paper by Silicon Graphics which was given out at the time and one from Liant which appeared afterwards if anyone is interested. Someone mentioned GKS - what a laugh that got! However as the speaker from Sun noted that it is the best selling application programmer interface and invited the audience to laugh while Sun cleaned up!
I found the panels on multi-media and on compression to be the most useful technically. The one on compression brought together some leading international experts in this field. Greg Wallace and Didier Le Gall were familiar names to me having read the special edition of Communications of the ACM in April 1991 which deals with compression (if you are looking for an overview this really is worth reading). Mussman and Girod made up the panel - 3 of the 4 panelists being European. I had become convinced during the exhibition and at other presentations that the JPEG and MPEG standards are going to be well taken up and this was born out by the comments made in this panel. Girod, not involved in the standards work as the others are but as researcher in the field, felt that the standards would win through against proprietary proposals but that research into model based compression was worth watching for significant leaps in the future.
The importance of compression standards and interchange standards was accepted at another panel on multi-media. The panelists were less keen on standards invading the area of user interfaces and stifling creativity. They were suppliers though so maybe that is not so surprising. It is through the interchange of ftles using standard techniques that is important.
The final panel I attended was on From Perception to Visualization. There was an opening quote from the Bible: where there is no vision the people perish (Proverbs XXIX, 18). The panelists were stating the need for good practice guides for the use of graphics. There is much research now on human vision and we really should be taking account of this. Issues such as motion perception and stereoscopic perception, being colour blind were presented. The importance of luminance was stressed. The fact that perception is affected by our knowledge of the world is also an issue. I suspect that AGOCG need to look at this as people are increasingly producing colour images to aid understanding. It is human understanding that we are trying to assist so we have to take human vision into account.
On balance, I did find Siggraph useful. The exhibition offered something which I could not get in the UK. The panels were good but like the papers and courses, I can go to many events in the UK and Europe which are at least as good (one issue is that it is probably cheaper to go to the US than Europe at the moment). It was so large though that I made very few new contacts. The party was so large everyone had a map showing where the various items of food were! I have never been to a party with a map before. I do not think many people seemed impressed with how crowded and uncomfortable it was.
Chicago - my kind of town? Yes, it was a good place to visit and I enjoyed the few days I had afterwards looking round, though the weather was poor. Any sport-nut like me must go to Nike Town on the main shopping street if you go to Chicago. Three floors of Nike kit and a sort of museum - Agassi's first set of white Wimbledon kit signed; Michael Jordan's boots. You can try on your new boots on a wooden floor and try shooting in them at the ring provided. Quite an experience to visit. And talking of sport and Frank Sinatra (we were?!) regrets? Yes I have one - that I was only in my bedroom when Linford Christie won the 100 metres and not in a bar so I could celebrate in public! NBC described him up to that point as just a footnote in running history! Get me back to the UK fast!!!
Following the previous two workshops: June 1990 in Rennes; and May 1991 in Barcelona, the Third Eurographics Workshop on Rendering was held at the University of Bristol from 17 to 20 May 1992. The special theme of this workshop was the use of parallel processing in rendering. The workshop attracted a large number of papers from 14 different countries. Of these, 20 papers were selected for presentation along with 3 invited papers. After a cheese and wine reception hosted by the Department of Computer Science on the Sunday evening, the workshop was opened by Professor Mike Rogers on Monday morning.
Each session of the workshop was devoted to a special topic: progressive refinement algorithms, ray tracing, monte carlo techniques, reconstruction of intensity functions, parallel algorithms, radiosity and sampling, physically based rendering, and, parallel radiosity algorithms. At the end of the day discussion sessions, pre-selected members of the audience gave their views on the papers for the day. The authors where then given an opportunity to respond, before opening the discussion to the audience. Needless-to-say, there were a number of lively discussions.
Derek Paddon's invited lecture dealt with the parallel graphics research carried out at Bristol, and concluded that the number of unsolved questions in parallel processing is likely to increase rather than decrease as technology advances. Paul Heckbert tackled the issue of discontinuity meshing for radiosity. He presented an algorithm for solving discontinuities such as shadow boundaries that places mesh boundaries directly along discontinuities. These algorithms offer the potential of faster and more accurate simulations. The final invited lecture was by Michael Cohen. In his entertaining lecture, he posed the question: Is image synthesis a solved problem? Included in his discussion was the results of what he called a very unscientific survey, in which he had collected the opinions to this question of a number of researchers in the field. Michael concluded that we still have our work cut out, not only in terms of research, but also in convincing the rest of the graphics community that there is still a great deal of valuable research to be done.
A trade exhibition was held concurrently with the workshop on the Monday. This culminated with a sherry reception and poster session in the evening. The conference dinner was a mediaeval banquet held on Tuesday evening at Caldicot castle in Wales, with Lord Pete and Lady Jean Shirley presiding. The 68 delegates attending the banquet were from 15 different countries. Large quantities of mead were consumed, with the international guests soon getting to grips with eating all courses with just a dagger.
The workshop once again proved the value of such small events in getting together researchers in an atmosphere which encourages lively discussion and interaction. We all look forward to the Fourth Workshop which is to be held in Paris next year.
Copies of the proceedings of the workshop are available for £25 including postage.
Apart from the obvious difference of size, SERC differs from other Research Councils in some important ways. The large majority of SERC funding and activity is directed to supporting research and studentships in universities and other higher education institutes (HEIs). About 15-18% of the SERC funding comes through its laboratories Daresbury Laboratory (DL), Rutherford Appleton Laboratory (RAL), and the two Royal Observatories.
This leads to a great diversity of computing environments with graphics being no exception. This article therefore makes no attempt to be comprehensive.
The balance of funding between HEIs and the laboratories means that one of the roles of the latter is to provide support for research in the former. The role of computer graphics in this research support is an important one.
Starlink provides astronomers throughout the UK with computers and software for the interactive analysis of astronomical data. Astronomers at over 20 sites in the UK have local access to a Starlink node and thus also to a common set of Starlink software and data.
An astronomer using Starlink is likely to come across computer graphics by using one of the many astronomical application packages in the Starlink Software Collection. For many there is no need for direct graphics programming.
For Starlink users who need to program, a number of graphics libraries exist. The user could choose to link their program with a library that provides a number of plotting techniques suitable for scientific programming. One such library is the NCAR library, from the U.S.A. National Center for Atmospheric Research.
Most of the Starlink application software is founded on a common set of graphics libraries at a more basic level. This includes the GKS standard, but also the SGS library (Simple Graphics System - based on GKS) which provides a simpler but still flexible graphics programming interface for the needs of people plotting data. Image display and interaction is also a basic requirement in astronomy and where this is not provided by the standard libraries, Starlink provide the IDI (Image Display Interface) library.
Particle physics is also a subject area where collaboration and sharing has taken place, although the community is more disparate than Starlink. The influence of software initiatives on a Europe-wide scale led by CERN is strong.
In recent years, the particle physics community has used PAW, a package from CERN, which closely couples data analysis and data plotting, a facility which has proved valuable in practice. Depending on the graphics system in the local environment, PAW can use GKS, GKS-3D or PHIGS, via a common intermediate layer.
At a basic level GKS is used and PHIGS is gaining acceptance because it is widely available on workstations.
The graphical needs of engineering researchers are perhaps more diverse. This is shown in the purchased application software, to some extent reflecting the different engineering disciplines, including mechanical engineering, electrical engineering, civil engineering, process engineering etc. Also the underlying needs vary: the graphical needs of someone engaged in interactive design may differ from the graphical needs of someone investigating phenomena which require the interactive exploration of data.
Graphical support is provided as part of the SERC's EASE programme, which issues the bi-monthly (Yellow bannered) Engineering Computing Newsletter. One approach is to focus on the basic graphics software. This includes RAL page (viewer for PostScript) and implementations of ISO standards which have been produced in cooperation with Central Computing at RAL, such as RAL-GKS and RALCGM. These graphics products are all available by network and have been described in recent issues of the Graphics Newsletter. PHIGS has been the subject of an evaluation at RAL and this has resulted in a CHEST dea1- this too has been described in recent issues of the Graphics Newsletter.
Beginning more recently, attention has been paid to the visua1isation of engineering data. The inauguration of SERC Engineering Board's Visualisation Community Club was announced last year. Although primarily intended for UK academic engineering researchers, its membership is open to anyone interested in visualisation either as a researcher, developer, user, or just plain trying to decide whether to be a user. The next event after publication of this newsletter is Visualisation Software to be held at RAL on 11 November. Particular concerns of the Community Club have been: evaluating visualisation software; getting data in and pictures out; building on visua1isation software by extending the range of techniques; providing ways for users to get started and solve problems.
RAL staff have played an active part in the definition of graphics standards, including GKS (currently undergoing a revision), CGM, PHIGS, PHIGS PLUS and language bindings and of the proposed image processing standard, in several cases acting as rapporteurs or document editors.
Daresbury is one of the four SERC laboratories and some of the trends reported there may be of interest elsewhere.
Cheap raster graphics and Unix workstations have revolutionised the way computer graphics is done at DL; compared with the previous batch-oriented approach, users inspect their graphics interactively and print only the interesting or publishable material - the traditional plotters are hardly ever used.
Availability of free software makes for a rich environment of graphics tools. However collecting and installing them centrally requires an allocated proportion of a person's time, to say nothing of users' DIY efforts which are uncountable.
CHEST has made a big difference. UNIRAS has been important and it is expected that AVS will be so in the future.
The application project teams (for example Synchrotron Radiation Source (SRS) teams) produce computer graphics which are novel and specific to their application. DL's central graphics provision is seen as enabling users to make their own choices about standards and free software and also about hardware.
Because of the diversity of graphics usage in SERC, this article has provided a few examples and initiatives known to and reported to the author. Acknowledgements are due to Paul Jeffreys (RAL) and Royd Whittington (DL) for their information and comments.
