advised Ferster that they were already considering another start-up in the same field and that it was a problem to go any further. Kaiser is still amazed to this day:
Who knows what would have happened if Bill Ferster had called me before Bill Warner?
Despite Greylock’s rejection Ferster realised that his main requirement for outside help was in sales and marketing. He organised for Pansophic to market the EMC product once it shipped. Ferster recalls:
You can't underestimate the value of the support from Jim Treleaven, Pansophic's General Manager for Graphics. He helped co-ordinate the office space and sales team for the new venture.
John Schwan had seen Ferster's business plan as his advocate within Pansophic.
We had a large sales team and it would be that group who would sell the EMC product when it was complete.
Ferster enlisted the support of friend Charles (Chuck) Rieger.
Bill Ferster and I had known each other for quite some time prior to our collaboration on the EMC editor, so the foundations were based on our mutual friendship and high degree of professional respect we shared for each other. My academic and professional background had been a mixture of AI and hardware and software design, while Bill's talents lay squarely in software design and architecture, with particular attention to user interface design
Ferster recalls:
The idea was to try and build a specialised appliance based on the PC like my previous product Cameraman but this was pre-Windows. We had to build our own graphical user interface.
Rieger continues:
Although my previous business partner and I had stopped further development on our SCION SoundSpace 32-track audio prototype it was still up and running, with a real-time UI that permitted multitrack recording and editing via up to 32 x 30MB SCSI disks running in parallel, each supervised by its own 80186-based controller card, represented as one of the tracks in the visual UI.
So the SoundSpace was the starting point for the EMC collaboration, since most of what SoundSpace did was almost identical to EMC's needs, i.e. managing real-time data streams to and from a hard disk in a scripted environment.
The collaboration proved to be a natural fit between Rieger’s hardware and embedded software knowledge and Ferster's vision for a nonlinear system. Soon after Editing Machines Corporation rented an 1889 former mansion in the Dupont Circle section of Washington D.C. Ferster recalls:
We developed the concept and the idea based on hard drives because it was easier but we always felt that the removable optical media was an important part of the equation. We knew people wanted to be able to transport their media around.
Ferster set a goal to create a system for the SMPTE show in October 1988.
LINK AND SUPERMAC
Meanwhile a third nonlinear editing company, and the only one based on the Macintosh, was struggling. In order to attract investors for their LINK editing system (above) and save money Don Kravits and Jerry Eisenberg rented a small one-room office in Studio City close to the main motion picture studios.
Kravits had settled on the Apple Macintosh SE computer as the platform to run the LINK because it offered several advantages not seen before.
Dan Knight recalls how the SE subtly broke the Mac mould:
From a practical standpoint, the SE's biggest improvement was a second internal drive bay, which could hold a second floppy or something new from Apple, an internal SCSI hard drive.
Following SuperMac’s success with licensing the DiskFit application from Dantz, Steve Edelman licensed the Pixel Paint program from Keith McGregor and Jerry Harris of Pixel Resources.
SuperMac revealed a sneak peek of the Pixel Paint software application. The ‘Mac Paint like’ program could be used on a standard color Mac II BUT SuperMac encouraged users to upgrade their Macs with a 256-color board. Edelman was one of the first to understand that third party hardware sold in greater numbers when bundled with software that empowered users. It was recipe he repeated.
PENCIL TEST
Galyn Susman grew up in Chicago and wanted to be a physicist or an astronaut. She told Emma Silvers in The Jewish News:
And then I went to Brown (University) as a physics major and found that I actually wasn’t very good at physics. So I chose something it turned out I was good at: computer science, computer graphics. Computer science really spoke to me and meshed with my way of thinking. I loved that way of looking at the world.
She joined Apple in 1986 where she worked on the development of the Apple ColorPicker for the Macintosh II and when that completed, Susman joined her Advanced Technology Group colleagues to make a grander plan::
In the fall of 1987 Apple's ATG decided that it was time to create a production quality animation for SIGGRAPH. The goal was to produce a piece of 3D character animation with high quality rendering. The challenge was to create this piece entirely on Apple computers, specifically on the Macintosh II
VP ATG Larry Tesler backgrounds the decision:
We wanted people to know the Macintosh was becoming a better and better machine for doing animation applications.
The ATG team also wanted to better understand what problems creative professionals may encounter with such a project and counter them with upcoming applications. In response Susman, Andrew Stanton and Nancy Tague created an animated movie called Pencil Test, a light hearted story where a pencil icon from a Macintosh drawing program came to life. In a prescient move, the team adopted an open approach to the challenge rather than build an integrated system.
While the latter would have given ATG greater control, Susman argued that the:
Open module systems offer one of the main advantages of Macintosh: integration. With the Macintosh consistency of user interface, different modules from different publishers have the familiar user interface. The best drawing program can be used with the best animation program while using the best video card.
The open versus integrated approach was one that was to repeat itself across the development of desktop video. A key element of their work was to use off-the-shelf programs like Super3D from Silicon Beach Software while the 3D models were animated using MacTwixt, a Macintosh version of the public-domain animation package Twixt, developed at the University of Ohio by Dr. Julian Gomez.
News of the group's work soon spread across Apple and other teams contributed expertise and equipment. Research scientists Steve Perlman, Carl Stone, Victor Tso, Dave Wilson, John Worthington, Larry Yaeger and Jay Fenton added to the project.
Bruce Leak from the Color QuickDraw team assisted in trialling WYSIWYG animation lighting and coloring. He shared with Hansen Hsu at The Computer History Museum:
Pencil Test was a big effort out of the Advanced Tech Group. We in system software were always behind. So I wondered how the research team could afford to have 20 people making an animated video. How could that be important?
After the timeline of scenes had been animated, Susman and the team discovered that Pencil Test would take 96 days to render and output to an external source if they used a single Mac. Ken Turkowski and the growing team hacked together a distributed rendering program that linked 28 Macintosh IIs together via Ethertalk to output to an imagesetter frame by frame, with each frame taking an average of 30 minutes to render.
Engineer Al Kossow recalls:
We grabbed Macs from other groups, from our own group and we brought in