SIGGRAPH 98 History Project
Vol.32 No.3 August 1998
SIGGRAPH and the SIGGRAPH Conference: The Early Days
Dr. Robin Williams of IBM was the fourth SIGGRAPH Chair, serving in 1976 and 1977. As he discusses in the following column, his administration drove SIGGRAPH’s modern era and set the direction that resulted in the society’s vigorous growth. It’s appropriate that his fascinating remarks become part of our 25th conference celebration record.
Dr. Williams is chairing a panel on SIGGRAPH’s past (scheduled to include other former SIGGRAPH Chairs: Dr. Jim George, Sam Matsa, Jon Meads and Dr. Steve Levine) at the CG Pioneers meeting at SIGGRAPH 98, on July 22, 1998.
— Carl Machover
SIGGRAPH is now a large, professional society with the world’s most impressive computer graphics conference. It wasn’t always this way. SIGGRAPH began as a small special interest group (one of the few in ACM) with no official conference. It grew dramatically into what it is now as a result of deliberate choices made in its early days.
This, briefly, is the story of how it happened.
Professor Andries van Dam (Andy), Brown University, and Sam Matsa, IBM, were the founders of SIGGRAPH in the mid-’60s. However back then SIGGRAPH was a Special Interest Committee (SIC) as opposed to a Special Interest Group (SIG). The difference was that SICs had their Chairs appointed by the ACM President (usually recommended by the soon-to-be Past Chair) while SIGs were self-governing bodies with officers elected by the group members. Sam was the SICGRAPH Chairman and Andy was the Secretary. SIGGRAPH was officially formed in 1969 and the first Chairman was Ed Devine (‘69-’71).
Jon Meads drafted our first set of bylaws and in doing so officially named us SIGGRAPH: the Special Interest Group on Computer Graphics and Interactive Techniques. The “and Interactive Techniques” was added very deliberately as there was a strong tendency to focus on output and “pretty pictures” for the society and the conference. A few people recognized the need for considering the human side of the equation and this was the means of expressing that interest. (There was no SIGCHI at that time.) In the early days, SICGRAPH published a bibliography to help people know what was being done in the graphics field (it helped me as a graduate student), and then started occasionally publishing a newsletter. Bob Heilman took over publications in 1971, and did an excellent job of attracting contributions and turning the newsletter and also the Computer Graphics quarterly into regular publications that were significant in attracting and keeping members for SIGGRAPH.
By 1972, SIGGRAPH supported conference sessions at the national ACM meetings and the spring and fall joint computer conferences. The society matters, policies and finances were discussed during evening meetings at these events. Occasionally small, specialized workshops were held too. The SIGGRAPH Computer Graphics bulletin became well established and was sent to all members on a quarterly basis. It contained a few papers (not formally reviewed) and general information about SIGGRAPH. It was the era when papers were being written for conferences on the topic of the past and future of computer graphics, ad nauseam, year after year, always predicting that next year computer graphics would come of age and really happen (while of course it was actually happening and growing rapidly all the time). Technology and facilities were still quite primitive and, for example, one talk by Bill Sass was severely hampered by lack of a slide projector in a hotel! Finances were still minuscule by today’s standards. In July 1971, for example, the society’s income was only $4,350.32 (compared with today’s income of more than $12 million), expenses were $3,989.21 (compared to today’s expenses of about $10.4 million) and dues were $4/year (compared to $85/year today). In June 1972, the balance was positive, but was only $1,409.31. There were some local chapters in Los Angeles, Syracuse NY, Washington D.C., and at one point Boston, and others were about to be started in Bologna, Italy and in Berlin, Germany.
Jon Meads became the Chairman (taking over from Ed Devine) and guided SIGGRAPH activities from ‘71-‘73. We were not yet a full self-governing body but had a board of advisors and key officers, but no elected directors. The members were industrial and academic experts, users, researchers and students with common interests, and the society’s purpose was to advance the state of the art of computer graphics and interactive techniques.
Bob Dunn became Chairman for ‘73-’75, and made the important observation and statement that SIGGRAPH was the “only organization that can effectively serve as a nationwide focus for computer graphics.” He drove SIGGRAPH to make this really become true. A membership survey at the time showed interests to be quite varied: covering simulation and modeling, text editing and composition, computer generated art, cartography and mapping, computer aided design and, of course, computer graphics software and hardware. Jim Foley was Vice Chairman for conference activities, and focused on promoting CG workshops and sessions. Many CG sessions were sponsored, for example at the International Joint Conference on Pattern Recognition, at the European Computing Congress, at Design Automation and at the ACM National conference.
In 1973, a most significant move was made by Bob Schiffman and Jon Meads — to organize our own SIGGRAPH conference, planned for Boulder, CO, in 1974. It occurred during a National Sciences Foundation Workshop on Computer Graphics in Boulder (in ’72 or ’73) when Bob and Jon decided that it had been too long since there had been a conference on computer graphics per se and decided to make one happen. Bob would be General Chair and Jon would be Program Chair. However, there was a problem. ACM was having some financial difficulties at that time and a prior SICGRAPH Conference on Computer Graphics in Medicine held in 1972 in Pittsburgh was a financial loss (although a technical success in spite of itself, which is another story). Bob and Jon felt the only way we could get approval for the conference was to run it so that ACM would not be financially liable. So Bob arranged that the conference would be jointly sponsored by the ACM and the University of Colorado. To keep the costs down and reduce risk, they decided not to print a proceedings but to have the conference papers printed in Pergammon Press’ Computer Graphics Journal which Bob was initiating (this is why you can’t find copies of the proceedings of our first conference).
We were now becoming large enough to think about focusing on just one good conference rather than the many sessions that we sponsored at other conferences and the focus shifted to this new opportunity. SIGGRAPH 74 thus became the first of our SIGGRAPH conferences. Around this time there was also a call for a computer graphics standard (a callable library), that was destined to become a significant development. A workshop was held on machine independent graphics and a Graphics Standards Planning Committee (GSPC) was formed with Karl Ryden (Chair) and about 21 members. Another development, run by Tom DeFanti, was to collect interesting films (back then a mathematical collection included, for example, two films on the Sine function, one on plotting and another on a vibrating string).
In ’74 there was yet another paper on why is computer graphics always a year away? People were getting good at perfecting papers on this topic.
SIGGRAPH 74, which was officially called the 1st Annual Conference on Computer Graphics and Interactive Techniques, took place in Boulder and was a great success. More than 600 people attended from many parts of the world, facilities were very good, and because it was held at the University of Colorado campus, many people stayed around for the evenings and continued to create an atmosphere of camaraderie. There was an introductory, two-day workshop to help people get up to speed so they could understand the main conference during the following three days. Working sessions for GSPC were held during the conference, and Larry Rossler became the new Chairman.
SIGGRAPH 75 followed the next year. It was held at Bowling Green State University in Ohio (David Fulton was Chairman, and Tony Lucido was Program Chairman). A key aim was to put on a conference with high quality papers, and indeed many were submitted and a few of the best ones were chosen for the technical sessions. A first Eurocomputer ‘75 conference that featured computer graphics was also held that year. Technically it was a time when raster graphics was just starting to become more popular (now, of course, it is the dominant technology).
In his outgoing remarks covering ‘73-‘75, Bob Dunn noted that we had now held two national conferences, three tutorial workshops, six other workshops and seminars and sessions at national ACM and computer conferences, had improved our publications (with more and more regular contributions), had started a more frequent newsletter, called SIGGRAPHITTI, and had started work on standards with GSPC. Membership had grown from 1,200 to 2,200 and our first elections were held. I was elected Chairman and three other officers and four directors were elected by the membership. We were growing and becoming a more and more organized and professional society.
SIGGRAPH 76, the third Annual Conference on Computer Graphics, Interactive Techniques and Image Processing, was held in July at The Wharton School, University of Pennsylvania. Tom Johnson was the General Chairman and U. Pooch was Program Chairman. This conference was the first to have a few specialized, formal equipment exhibits (but no list was published). Overall it was good, with some tutorials added to the main conference sessions, but the original excitement was not there for some reason. We could tell that to grow and to be known as the key national computer graphics conference, we would have to radically change the format for SIGGRAPH 77 and beyond. The ’74 and ’75 conferences had made surpluses of about $4,000 and $5,000, but in ’76 the surplus fell to $1,000. Because the conference helped subsidize membership fees and other activities, it could have become a problem.
During the ’76 conference, GSPC, which was one of our key activities and had languished in the previous year, was reactivated. Bob Dunn and Bert Herzog became the leaders of a group of 25 expert volunteers, divided into two groups to deal with short-term issues and longer-term core issues. There were other standards efforts going on elsewhere: GPGS in Holland and Norway, GINO-F in the U.K., and there was a desire to collaborate with GSPC in the U.S. Because of the renewed interest and a real promise to make headway, some precious SIGGRAPH funds were allocated to this renewed activity. As the activity picked up and became a serious effort, more funds, finally growing to $21,000, were spent by SIGGRAPH (‘76 - ‘77) for GSPC people’s telephone calls, meeting expenses, computer terminal lines and publications. Although the groups met at one another’s homes to keep costs down, it was a large expense for those days and put more pressure on us to have a much expanded conference to help fund such desirable technical activities without having to raise dues too much.
Technology in ’76 was changing rapidly. Raster graphics was competing with storage tube displays and vector graphics and color was becoming more popular. Nevertheless you may be amazed to know that we still had papers on topics like: 1) a gray-scale character set for the production of pictures at a printing terminal; and 2) curve sketching on a Teletype. Meanwhile a second workshop on Methodology in Computer Graphics, run by Richard Guedj, was held in Seillac in France (considerably advancing the standards efforts). We sponsored a senior person to attend the IFIP workshop on Recommendations on Methodology in Computer Graphics on the condition he write a report to members in the CG bulletin, (he did).
Plans were made to hold a revitalized SIGGRAPH 77 conference in San Jose, CA. It was decided to move to a major hotel (a Hyatt) with excellent A/V equipment, exhibition space and large meeting rooms. We planned to invite industrial companies to buy exhibit space and for us to aggressively promote and sell it. Steve Levine was the General Chairman, Jim George the Program Chairman and with me, we formed an executive conference committee that met frequently to pull it off. We hasten to add that there was a large group of people planning, supporting and running the conference (a committee of about 15 people) and we had the full support of the rest of the SIGGRAPH board. A strict reviewing and referee process for submitted papers was started to aim for the highest quality.
Norm Sondak was the Publicity Chairman and the conference was well publicized, especially in the CG bulletin, and emphasis shifted away from sponsoring other activities to concentrating on the SIGGRAPH conference and the GSPC standards activity. Fortunately it all paid off, and the results were very gratifying and justified our decisions. For the technical program, 140 abstracts and papers were submitted and about 40 were selected for the conference, providing the high quality that had been sought. They were published on quality paper, with color graphics, giving a fine professional appearance to the conference proceedings.
The results of the standards activities were presented at the conference also and published for all members in a special edition of the CG bulletin. For the exhibits, 38 vendors including the best in the world at that time, showed their most advanced technology and products, and it was exciting. There were two workshops and films and videotapes were shown over the hotel TV. A total of 750 people registered for the conference and many positive comments about the changes were received. One person (in later years) said it had changed his professional life. Andy van Dam said it was like night and day in comparison to SIGGRAPH 76.
When the conference was all over, SIGGRAPH was still in the black financially, an extremely important point, because we had taken a big gamble in all respects, technically, socially, visibly and financially. If it had not succeeded, SIGGRAPH might well have reverted to a small professional society again. ACM would have had to put more controls on the SIGGRAPH board, and rightly so. The board had pushed aggressively forward on several fronts, controlling our own finances, making rapid decisions and submitting paperwork later, changing the type of the conference by increasing the number of exhibitors, changing the quality of the publications (at considerable expense) and by starting a refereed section of the CG bulletin and publishing algorithms. With more than 3,000 members, SIGGRAPH now became even better organized, published its budgets and became strongly established.
As often happens, success begat future success, and the SIGGRAPH conferences became larger and larger in subsequent years. They quickly grew bigger than any other ACM SIG’s conferences, and SIGGRAPH soon put on a much larger conference than ACM itself. The number and size of the exhibits also grew very large and required much larger facilities, so much so that only a few sites could actually host a SIGGRAPH conference any more. During the ‘80s, reservations had to be made years in advance for each future conference. Because of this growth, the SIGGRAPH board had to seek paid contractors and professional help to organize and run the whole conference (it became far too much for volunteers to do). For example, by 1997 the total attendance grew to 48,700 overall (compared to just 600 overall for our first conference in 1974)!
But I’m getting ahead of myself. In 1978, the IEEE Computer magazine published a report: interactive graphics poised for takeoff. Yet again it was suggested that computer graphics was about to become of age!
Jon Meads helped me with some of the early history.