Timeline Analog 5. John Buck
to the Toaster, plus the a/b edit controller, built in TBC’s and well everything, short of a second VTR, that was needed to do A/B roll post.
WOW, yet the price of 15k, a big drop over 200k, was still a high price for a small shop. How I drooled over the ads…oh to dissolve between two live sources, oh to have a DVE.
Pinnacle saw an opportunity to make a complete system. Co-founder Ajay Chopra looked at the options.
We were the absolute leaders in real time special effects in the On-line video editing market, but then along came digital nonlinear editing. We could see that it would impact the traditional editing bays and we felt that we already had this core real time technology, which is what the market wanted and liked, but perhaps we should look at creating our own digital editor and wrap it around our real time product.
We had been in the market long enough to understand both editing and effects. And to set the scene, the traditional editing set-up was changing with Avid and EMC arriving but in those days editing was also much more simplistic than it is now. Simple A/B editing without multilayering and compositing.
We believed that it was feasible to add our own editing application to a hardware card because you could argue that the special effects of that era were harder to do and we had already done that.
Instead of building a package from scratch, Pinnacle decided to buy an existing player. It courted the clear number two in the editing market, Bill Ferster’s Editing Machines Corporation (EMC) in Washington.
While Avid used a Mac platform, EMC had opted for the platform that Pinnacle knew intimately, the PC.
Chopra planned to mesh EMC's software with its hardware and grow Pinnacle into a video equipment market with many products. Talks failed to progress and were then abandoned.
FAST
In March 1992 Markus Weber finished his PhD and was looking for a job in a tough employment market. He landed an interview at FAST Electronik on Landsberger Strasse in Munich.
I managed to get an interview with Matthias Zahn's second in charge, Ali Adelstein the next evening, followed by signing a contract half an hour later. Matthias and Ali had been hanging out with a mutual friend, who was a cameraman and editor, a couple of weeks earlier.
Bradley Giotes adds:
Martin Leckert, a friend of Matthias, was a professional camera operator in Germany. He bought some consumer video equipment to try to do his own editing at home and became very frustrated. He called Matthias and begged him to create something better.
Fast had sold more than 10,000 Screen Machine digitizer cards in Europe. Weber continues:
They were talking about ScreenMachine and they came across the idea of sticking two ScreenMachines into a computer, hook them up to another chip on a third board, mix the two images of the frame grabbers together and feed the result to a Video Out.
Matthias did a rudimentary drawing on a napkin, and the next day set out to put the idea to the test. This was the birthday of VideoMachine.
The FAST team managed to put together a basic VideoMachine (VM) prototype to demonstrate at CeBit, Europe’s biggest consumer electronics show. Weber recalls the opinions they received.
People wanted an entire application that would replicate an online video editing system on a PC for a fraction of the cost of say a traditional Sony BVE-9000 or CMX-3400. When the team got back to Munich from the CeBit show in Hanover, all hell started to break loose.
Everyone recognized that here was a really huge business opportunity if we could pull clear in the editing app, and this signalled the instant death of my job building the database. You could not expect to go to market without offering some kind of device control to drive two playback VTRs and a third recorder, essentially an Edit Controller.
They needed someone who could be expected to properly deal and model the complex dynamics and ballistics of an intricate mechanical system called a VTR, and devise a system to sync them to an output device. To be clear: no one had any knowledge of video technology beyond what was required for a frame grabber.
To call us video virgins would essentially be overstating the fact. We had absolutely no clue what we were getting ourselves into.
Weber was charged with coming up with a software solution to the edit controller issue, while the rest of the team was setting out to develop an editing system on top of this. He recalls:
Everyone was charged on adrenaline and went into hyperactive mode.
The CeBit reaction convinced Matthias Zahn to continue to the next level, but on two platforms. He created a team for the PC development, and partnered with another Cinetic of Karlsruhe to make a Macintosh version.
Weber recalls:
Both FAST and Cinetic were deeply rooted in the German market, and then the European market. This was the home-base, where the money came from. The Mac was a definite no-show in Europe and mostly still is.
On the other side, the market was comparatively small, when compared to the potential the US represented. So everyone in Europe knew exactly where the future was, and everyone was struggling to get into this market.
This market was dominantly Mac at the time, with the PC a distant runner-up. If you wanted to work both markets successfully, you needed a solution that could survive in both markets, Europe and the US, and this, in turn, meant having an ISA and a NuBus solution.
Of course, it also meant, to close the circle, that in Europe the emphasis was showing that things worked on ISA (you could show the Mac, but no-one really cared), while in the US you had to showcase the Mac.
FAST had three deadlines to meet for the Video Machine.
Ali Adelstein, Markus Duerr, Michael Berneis, Christian Rogg, Martin Regen and Weber needed to prepare for spring COMDEX, followed by the International Broadcasting Convention (IBC) in Amsterdam and then NAB in Las Vegas. Weber continues:
VideoMachine was kind of a super strange beast in more than one aspect and Fast was trying to cover all bases for an uncertain future. In 1992 the PC race was still more than open, and the Mac seemed to be far better suited to video processing than the Windows 3.1 PC and it was.
As it was unclear who was going to win, Fast developed a hardware architecture that was adapted to two platforms: The ISA bus on the PC and the NuBus on the Mac. That was a rather drastic difference in terms of what could be done in principle.
The PC version used the ISA bus concept, which in 1992 was already ageing. It was either 8 or 16-bit wide, ran at 8MHz and typical average transfer rates achievable were 1-2MB/s.
EISA was just about to be introduced, but it was physically and developmentally much more demanding and did not promise to be a successful contender compared to ISA (much higher costs, little revenue benefit), because the innovation-advantage was not perceived as large enough to completely redevelop a hardware solution.
The situation on the PC-side would essentially not change until PCI was introduced. The NuBus, by comparison, already sported a 32-bit interface that could average 10 to 20MB/s throughput in it's original version (10MHz). To be clear, that was about 10 times what ISA could do. So the playing field was anything but level.
For the PC version this meant that over the lifetime of VideoMachine, development was always at least one step