Читать книгу Timeline Analog 3 - John Buck - Страница 24

Ron Barker’s quest for seed money lead to many dead end interviews with potential 'angels' before he stumbled upon the Boston based scientist and engineer Jean Tariot. Tariot had recently sold Incoterm, the computer company he had started with Maurice Upton.

Jean Tariot was in the middle of a holiday in France when his name was suggested to us, so we had to wait patiently until he returned to pitch the system but it was worth it. He loved the idea so he contributed $60,000 for 20% of the company. Tariot not only put in money but with his technology experience he told us what was missing from our business plans and financial forecasting. And cleaned up my act for me!

Schuler recalls:

He was intrigued enough with our concept to provide the initial funding of $60k, that we needed. He also became the chairman of our board and was primarily responsible for securing the full funding that allowed us to bring our concept to fruition.

Tariot introduced Barker to William (Bill) Field from Prudential Insurance Company of America who had previously bankrolled Tariot’s Incoterm. Field invited Ron Barker to pitch the idea to him.

I told Bill Field that for $1m, I would prove the system worked and get ten initial orders within 12 months of us starting. After a one hour pitch he said: "I will take all but not less than half". He invested and promised us another $1.5m if we hit the target.

The new company to build a next generation editing device was formed.

The Video Composition Corporation (VCC) received its first round of funding in mid-November 1982. Ron Barker recalls the enormity of the task:

We had $1m and a goal to make a working prototype and close ten sales in 12 months. That was the trigger for more funds to build a new visual editing system. Before there was the Mac, before DOS and certainly way before affordable hard drives. I asked Chet if he could get us a deal on a Masscomp computer because of his association with them and we would use that to create an editing simulator.

Barker and Schuler then settled on a plan of action. It was a list that dominated their lives for the next year, laid out in point form about what their editing system (previous page) should be capable of.

*Multiple picture frame [image] presentation at all times including past, present and subsequent shot/scene transitions with picture [image] used as the frame label for all decision making.

*Continuous picture [video sequence] looping for repetitive transition or scene presentation with real-time dynamic control of transition points and form [type] of each transition.

*Hands-on control wheels that allow all [editing] decisions to be made without moving the hands or looking away from the picture [video sequence] display of the material being composed.

*Automatic transition, shot, scene and story Assembly with total flexibility to insert, add, or delete material at any time and rapid [single] step sequencing both forward and backward by transition, shot, scene or [complete] story [segments].

*Hard copy Storyboard print out of each composition for narration, re-composition, or final conformation by hand.

*Machine-readable decision disk for remote video [on-line] conformation.

*Finished programs or work prints recorded to industry tape formats (¾", 1" and 2")

Then the two men looked to hire key personnel as Barker recalls:

I contacted Ken Kiesel who I had known as a fellow actor at Vokes Theatre in Wayland and convinced him to leave Polaroid. Then I contacted Ed Moxon, who I knew from when he worked for a BTX competitor, to start the team off.

Barker knew from his BTX experience that professional tape machines were too expensive and inflexible to use in the multiple machine array that was planned to provide the instant preview so he purchased two Betamax machines.

Ed Moxon conducted a thorough examination of the VCR’s construction and schematics, and he determined that the Sony machines could be retrofitted with a microprocessor board of his design to accomplish what was needed.

Ex-Polaroid engineer Ken Kiesel joined Moxon in the design and development group.

When I first interviewed with Ron and Chet and they described what they were doing, my only reservation was the use of these consumer video machines. Talking to Ed Moxon convinced me that success was possible. Without his being on the team, even with the enthusiasm I felt for the concept, I wouldn't have left Polaroid.

Chet Schuler’s friend and former Incoterm engineer, Bill Westland became the chief software engineer.

I had worked on various uninteresting projects after leaving Incoterm, until I received a call from Jean Tariot. He asked if I would be interested in going to work for an exciting new startup company he was helping to get financed. A few days later, I met Ron Barker and Chet Schuler for the first time at the Red Coach Grill in Wayland Massachusetts and then we went over to Ron's house in Weston so they could "sell" me their plan. I didn't know anything about the film or video industries, but it took less than an hour for me to know that this was something special and that it would be a great opportunity for me.

The staff list at Video Composition Corporation on Domino Drive in West Concord, Massachusetts grew when Margaret Marsden, Ray Marchant, Michael Tindell, Michael (Mike) Lowe and Chuck Urian joined the group.

Ron and Chet approached me to consult with Montage. They wanted someone with my background (broadcast & cable) to specify potential “end-users”. I also helped with some of the design flaws that would prevent intuitive use of the system by the film/tape editors around the world.

Tindell was coming straight from an MIT Computer Science and Engineering degree. Schuler adds:

Bill Westland headed up the software development for us because his experience was in real time programming, controlling machines and not just databases. We knew we needed someone who could work on tasks other than writing great code and manipulating data. Mike Tindell worked in 'C' and he would hear you saying 'we should try this as a function in the system' and then he would just disappear and type as fast as he could and then come back after five minutes ask, "How does that look?” He was an amazing programmer; great instinct and just knew it inside out.

Schuler remembers that the group had a deadline but he wasn’t about to be hurried into a mistake.

A lot of people don’t realise that you can save yourselves a lot of time if you're not in too much of a hurry. We spent more than a month after we got the money, not working, just planning what we were going to do. We laid out our plans for what we wanted to accomplish in general terms and our criteria for achieving our goals.

We were treading on pretty new ground and because of that we wanted everyone on the team to know where we were going. Since we needed to do both hardware & software development concurrently, it was necessary to have a very close knit group with each member contributing to and constantly aware of the progress of all of the parallel paths of development.

That isn't easy because if one isn't done, the other can't be tested and so on.

Together we mapped a general outline for the overall design of each aspect of the project. Because most of our specific hardware requirements and software requirements had no precedent, we often had to choose between somewhat untested and untried alternatives, ahead of the actual product design decisions.

Bill Westland recalls an early glitch:

When we first started calling computer equipment vendors for information, Ron prohibited us from using the full name of the company because he was afraid people would find out what we were building and he wanted absolute secrecy. So, instead of saying Video Composition Corporation we told people our company name was VC Corporation.

One of the vendors I spoke with was uncomfortable with that name, because he was a veteran of the Vietnam War and for him VC was a dirty word.

Ron Barker recalls:

The new name Montage came after the funding from an employee suggestion box. Mike Lowe was the originator of the name Montage. I drew up the logo by cutting off the lower portions of the letters. The newly christened Montage team had a collection of equipment befitting a start-up. A Masscomp computer, Schuler’s S100 with a 20 MB hard drive, a dual 8” floppy drive, a 5 ¼” floppy drive, a video frame grabber, a couple of Betamax decks and a real time relay control board. Ken Kiesel recalls the state of the company: We had desks and telephones, one computer, no parts bins, test equipment, or even supplier catalogues.

Despite their humble beginnings, Schuler recalls the confidence of the group.

The key idea was the concept of not just using pictures to represent timecode locations but manipulating the edit in such a fashion that it wasn't even something you would think twice about as the editor. How we would place them and move them around would make or break the device.

This was a very early application of cut and paste editing that it so familiar today, picking up a picture and dropping it down again like a virtual video script. We needed to work out a way to get it done, what we needed to do, in spite of the technology. Hanging over our heads was that we needed to use existing technologies. I already had a digitizer, so we used that early on.

We could control it from the Z80 processor and the digitizer would ingest the video and audio and timecode so that it could be used by the systems but it ingested only in a temporary fashion.

In other words, unlike today's systems where memory and processors are cheap and capacious, we didn't digitize and keep everything. Instead we opted for a method where we just digitized the frames around where the person was editing and when the editor had completed the editing and saved the edit decisions, they moved on and digitised a bunch of new scenes

For stage one of development, the team split into two informal groups. Westland and his colleagues set up a newly acquired VAX computer and began the task of writing and testing code for the editing system while Ken Kiesel focused on a full size mock-up of the intended system so Ron Barker could get feedback from prospective clients. Schuler recalls:

The (overall engineering) strategy that we settled on to help us avoid investing too much time in dead end ideas was to have Ken Kiesel run ahead of the actual product design team by creating the working simulator. It was a mixture of readily available standard computer and hardware along with his ingenuity in how to test alternative approaches.

It was more important especially in the early days to see how it played to the user and whether we could accomplish what we wanted with the way images were done.

It didn't need to be right, just enough to test and see if worked or if we needed to change our approach. Ken worked on a lot of things ahead of time that weren't quite ready to be incorporated into the final product and didn't have to be, just so we could test out our concepts - so they made sense. It saved us going down any blind alleys for weeks on end because we could quickly assess the value of our ideas.

It was a small team but if I had more people, it would have taken twice as long!

Kiesel recalls:

My first assignment was to built a simulator for the system and have it capable within 3 months of performing simple cuts and splices of material from a pair of laser disk players, including the user interface with scroll wheels and buttons and 14 monitors showing picture labels. I immediately realized that I couldn’t possibly accomplish this task alone.

I convinced Ron that we needed a mechanical engineer to design the actual controls, or we’d wind up with a wooden contraption that would give the operator splinters every time he pushed a “button”. A contractor was hired within a couple of days and he quickly developed a panel with heavy brass detented wheels that felt really luxurious when you turned them.

Kiesel then convinced Barker to hire an extra contract programmer to help him make the deadline for completing the Simulator.

We used Chet’s Masscomp computer with five memory boards of 1MByte each to hold picture labels. It was a Multibus II computer and Chet had a frame grabber board for it already.

Bill Westland recalls:

We chose the Multibus II over the VME bus, even though the VME bus had bigger boards, because the Masscomp computer used Multibus and that allowed us to use the same screen display board for the Simulator and for the real product.

Kiesel continues:

We only needed a display board with 14 video outputs for the picture labels.

Schuler commissioned engineer Gil Drake a highly regarded engineer who had worked as a circuit designer at Dennison Paper Company, to develop a board as Kiesel recalls:

Gil Drake was a brilliant engineer and he knew video really well. Gil came to work for us along with his two sons.

Barker recalls:

We rented some space in and built a preview suite that became nicknamed ‘The Simulator’. It was essentially a representation of what we intended to build before we actually did! We literally built a mock-up in one room with white foam board and drew or cut out the video screens and the pictures on the screens and the wheels and knobs and control levers.

Then I went out and bought a Philips videodisc player and got a couple of discs burnt with demo material on them. We put the computer and the disc player in another adjacent room away from view.

Kiesel continues:

At that time my software experience was limited to assembler and I learned to my dismay that I would have to use C with the Masscomp. There simply wasn’t time for me to do this and get a real time simulation running with the time left and I was getting quite panicked.

Barker recalls:

Then we sat down and worked out a way for the computer to play a sequence of events that would look like what would happen, as the machine in the room would respond to my commands. It helped prove the sexiness of the approach without solving any of the problems.

With the 'Simulator' complete, Barker invited post-production industry leaders from across the United States to view the mock up. Everyone who came to the room dominated with dark green carpet on the ceiling and floors were under strict non-disclosure agreements. Barker explains:

Of course the whole idea was actually to show people and get them to tell as many people as possible about our system! We knew if we told them not to talk about it, that was the best way to get them to talk about it.

Away from the ‘Simulator’, the Montage engineers were able to experiment with building the real system. On the software side they dealt with VCR machine control, tested different sized gray scale images for legibility, tested labels to represent time code locations and printed sample storyboards.

For hardware design they had to confirm control lever, wheel and button user interface requirements. Schuler recalls the first of the hurdles.

The biggest problems we were to face weren’t the digitising or storing of the data but displaying it for the editor. How could we show the video rushes on displays or monitors? There were no multiple 'windows' on a single screen like you have with computers nowadays.

The Montage team proceeded with prototyping while Barker and Schuler debated the interface.

I remember that our investors Prudential continually harped on about the need for a keyboard. They would get through to Chet and then Chet would come into my office and pitch the idea of including a keyboard. And as far as I was concerned the only Montage people who were ever going to have a keyboard were Bill Westland and Mike Tindell, who were writing the code. End of story.

Bill Westland recalls that even the work at the keyboards was causing stress.

I guess the work that Ken had done with the ‘Simulator’ was more obvious in the early days and when the focus came onto us, it looked like we hadn’t done much for five months. Then over a period of a few weeks we made great progress and suddenly we were up with or ahead of the Simulator team.

All the while, the Montage team were concerned with Lucasfilm's status.