Читать книгу Timeline Analog 6 - John Buck - Страница 13

The relationship with KeyGrip’s technology partners soured.

Microsoft had promised to deliver ActiveMovie as the enabling Windows API and Truevision was to be the preferred hardware supplier for Wintel machines. After its launch in March there wasn’t another press release all year.

The Macromedia team received a phone call about the development of a Truevision driver. Phil Schiller told the subsequent U.S. Justice Department’s case against Microsoft:

The team from Truevision called my team at Macromedia and said, “We can’t deliver you your QuickTime driver for our card on Windows because we have entered into an agreement with Microsoft. And part of that agreement requires us not to make a QuickTime driver for our card for Windows.

Apple’s Dr Avie Tevanian added in his deposition:

Macromedia was working with Truevision to develop a ‘driver’ to ensure that its video capture card was compatible with (KeyGrip) Final Cut software. When Macromedia realized that Microsoft was not going to deliver ActiveMovie as promised, Macromedia decided to alter the development of Final Cut so that it would be based on QuickTime for both the Mac and Windows operating systems.

Although this work continued, Truevision eventually informed Macromedia that it could no longer continue with the development of a Windows QuickTime driver to support Final Cut. Mr Schiller was informed that Microsoft had agreed to invest in Truevision, but only on the understanding that Truevision could not deliver or support QuickTime drivers for the Windows operating system with Truevision products.

According to Mr Schiller, Truevision and Microsoft reached an arrangement that allowed Truevision to create a QuickTime driver for its video capture card. However, the driver could work only with KeyGrip and Truevision could not market, brand or refer to the driver as a QuickTime driver.

Despite negotiations with Truevision to resolve the issues, Macromedia management was now forced to adopt QuickTime as its API for both PC and Mac. It also needed to settle on a new hardware partner.

In August 1996 Macromedia had $100 million cash in the bank but with many businesses moving their focus to the Internet, its core business of making software tools for developing CD-ROMs was slowing.

The woes at Apple had also sent many Macromedia clients to the seemingly more stable Windows/DOS platform.

Chairman Bud Colligan believed Macromedia needed a fresh, outside perspective to survive and he replaced himself with Robert Burgess, former VP Software at Silicon Graphics.

Burgess had stated publicly that a ‘hard’ platform like CD-ROM was going to give way to a ‘soft’ platform like the Internet. The depth of the Internet as a platform was still unknown but Burgess was convinced that it could save his company. He told Fortune:

Basically, we bet on the idea that the Web could be a better place and that we could build some stuff to help it along. We really understood interactive media, and the products we created hit the sweet spots of the markets we targeted.

Burgess wanted customers to come to Macromedia for a complete website building package and it was unclear what role KeyGrip played in that mix.

The AliBaba team still needed to solve VCR control for its incomplete consumer-editing package. VP New Development, Bill Loesch recalls:

I had sold Mark Sanders on the idea of this project called AliBaba, which was an offline then online editing system. I had looked around at what others were doing and I had used Gold Disk's Video Director product myself. Video Director was essentially an off-line conforming tool (with no front-end non-linear editor).

It controlled the source camcorder (using LANC, Cntrl-M or IR) and the record VCR (using IR). It was cuts-only. As I got deeper into the detail of what Pinnacle needed to create it was obvious that one of the key pieces of technology that we needed was low-end IR machine control. I figured we could clone this technology and create a new front end.

Out of nowhere Loesch took a lucky break.

The computer publisher Gold Disk had sold more than 100,000 copies of its $99 video editing product called Video Director. Company cofounder Kailash Ambwani told the press:

Gold Disk has worked hard over the past two years to provide customers with affordable, easy-to-use software tools for enhancing their home video collection, with great success.

Despite the success Ambwani changed direction and focused Gold Disk on the emerging web conferencing market. Loesch continues:

Gold Disk approached us to see if we wanted to buy Video Director. It was Pinnacle's first acquisition after the IPO and we really agonized over that purchase!

Pinnacle's Mark Sanders knew the acquisition solved AliBaba's immediate problems.

They (Gold Disk) had this great deck control technology that we could use straight away. VHS tape machines are pretty basic, they are shaky and the ballistics change from one end of the tape to another. A 120-minute tape acts differently to a 3-hour tape and the existing Pinnacle consumer editing system was essentially predictive.

It would predict if you pushed the button on your VHS deck how long the deck would take to roll up and get to speed and lock. So you could easily be off five frames without batting an eye. It was very difficult and very frustrating for the home user.

Bill Loesch had the pieces to let home enthusiasts make professional-looking video.

We now had all the bits and pieces. Our basic software, an external ‘mixing’ box based on the Sherman ASIC from within Pinnacle, and the Video Director package from Gold Disk. Using all of the components the team came up with a new product based on the Video Director UI that allowed title overlay and fixed-frame transitions.

Ivan Maltz explains:

They (Gold Disk) had developed not only a tape-to-tape software program, but also this great and simple to use 'smart cable' with which you could control Sony Camcorders using their LanC interface and an IR blaster for controlling VHS decks.

Looking back it was a little Rube Goldberg but it worked! We built a new UI, replacing the Gold Disk program. Keith did some great work there to create a 1 - 2 - 3 editing interface, capture, edit, play to tape. We had started in 96 with nothing and by the end of the year we had a product ready to go.

Loesch concludes:

By acquiring their technology and several key hires, we certainly shortened the time frame for delivering the new product. But of course we also picked up substantial extra costs that we hadn't budgeted for. When you suddenly have extra employees you need to start shipping something much quicker than you planned for.

With the acquisition of Video Director, Chris Zamara joined Maltz, Thomson, Lane and former Abekas engineer, Jon McGowan. Maltz recalls:

The five of us formed the Studio development team for a long time, eventually cranking out Studio 1.0. This version of Studio used a custom external hardware box (plugged into the parallel port!) to capture video frames.

It also allowed us to more accurately control IR-driven camcorders using a "terrain mapping" algorithm for which we got a patent (#5917990). Still, it wasn't very accurate, especially with cheaper camcorders. The future obviously had to be all digital!

Mark Sanders recalls the decision to move.

We needed to get out from underneath this analog VHS tape editing system and go digital.

Meanwhile CBS was evaluating nonlinear editing equipment, and VP Don DeCesare told the press.

Where performance is concerned, Avid, Lightworks, D-Vision and ImMix are on equal footing, but all have drawbacks. D-Vision's system keeps crashing, and the Lightworks and ImMIX Systems are too much like film-editing machines.

Slow digitisation is shared by all four. Instability at Apple Computer is some cause for concern about the Mac-based ImMix and Avid systems.

Despite the technical issues that CBS was confronting, the very choice of participants in its shoot-out was telling.

In days gone by it would have been Ampex and RCA, or Ampex and IVC but now it was fight between companies, none more than nine years old.

After 20 years in business, Apple had shipped 25 million Macs. Much of its success had been built on the print desktop publishing revolution but to date it had failed to deliver similar results with video.

It had commissioned Avid to create a simple home video editing package that it could bundle with its new Performa 6400/200 VEE (Video Editing Edition) computer.

Avid Cinema 1.0 arrived at MacWorld in Boston.

The standard Mac hardware, had been changed to cope with video editing and included a PCI video capture card, 2.4Gb hard drive, an additional PCI slot, an integrated subwoofer for audio output and Avid’s software.

Avid Cinema is the first computerized video-editing system designed for the camcorder crowd, not professionals.

The Cinema interface used tabbed folders to direct home based editors through a template approach. The Step One tab encouraged users to create a story outline, clicking on the Step Two tab instructed the user how to import the video clip from a camera or VCR using the 'Video In' function that opened the Performa's TV/Video port.

There were 21 storyboard templates from birthday celebrations to graduations that instructed users on what kind of shots were needed.

The finished movie could be published to videotape using the 'Video Out' function that was linked to the Avid PCI card. More likely users would create a CD-ROM or QuickTime movie of their final work for presentations or web sites.

At MacWorld, Apple demonstrated a future release of Avid Cinema that used a 1394 (FireWire) card connected to a Sony camcorder for ingesting video. Apple luminary Jef Raskin reviewed the Cinema for Mac Home Journal:

With Avid Cinema, you can turn out videos chock full of sound, video effects, and flawless editing. This software/hardware package fall somewhere between amazing and incredible.

Avid's product line now stretched from the home to Hollywood.

The KeyGrip (above) development team moved into new premises at Redwood Shores Parkway.

William (Will) Stein joined as Director of Engineering.

Stein previously managed Apple’s MacOS Printing Group, followed by a stint on QuickTime.

I left Apple, and went to Macromedia to manage the xRes project that had been acquired from Fauve Software. Macromedia was making a strong push into the "content editor" space, with projects either shipping or under development for digital photo editing and paint (xRes), Freehand (Illustration), 3D authoring (Extreme 3D), a new audio editor from the DECK acquisition and of course KeyGrip in the video space.

Norm Meyrowitz, president of Macromedia Products, asked Stein to look at KeyGrip.

He wanted me to try to help the group "push the product out the door."

Stein took charge of engineering management and let Randy Ubillos focus on coding. Ubillos recalls:

Rob Burgess knew we were doing good stuff but not making good progress so with a better distribution of responsibilities, we moved forward. At that point KeyGrip had been built and shown to run on either the PC or Mac platform because we had done a lot of work to abstract the editing system from the playback system, a lot of that was achieved of course by using QuickTime.

As for the feature set, the list was always changing and evolving because the hardware around us was always evolving.

Stein continues:

When I started with the team, the project had been going on for a year and a half, and the executive management thought was that it was within 3-6 months of being ready to ship. I worked with the team through a schedule scrub, a very detailed review of all the tasks remaining to complete.

Tim Myers recalls the change.

It was really good for Randy that Will came in and managed the team. It freed Randy up to do what he is so good at doing. Right about then the pressure was increasing to deliver the product and Will helped us with a plan to get finished and to see the light at the end of the tunnel.

Macromedia held its User Conference in San Francisco and Apple’s COO Marco Landi announced QuickTime for Windows 95 and NT:

This QuickTime development project marks the first step in Apple's plans to move all of its core interactive multimedia technologies, something we call the QuickTime Media Layer, to key industry operating systems such as Windows, OS/2, and Unix.

He announced major editing features would be delivered in QuickTime 2.5 including the new M-JPEG file format, which allowed video editors to work with files independently of the hardware originally used to capture the media. Geoff Duncan wrote for TidBITS:

… unlike MPEG, it (M-JPEG) retains information about every frame of a movie (rather than interpolating between key frames), which makes it more suited to high-end production. QuickTime 2.5 will enable high-end video people to exchange and work more easily with M-JPEG video, provided hardware vendors adopt Apple's new formats.

Adobe, Intergraph, Equilibrium, Macromedia, Media 100 and Truevision were developing products to work with QuickTime for Windows and NT.

Engineers from all eight companies will be collaborating with Apple to architect key elements of the QT 2.5-Win product, including video frame grabbing, NT multi-processor support, memory management, media compression, and decompression as well as the high-performance playback necessary to ensure broadcast and professional adoption.

While Truevision announced support for QuickTime, it had been prohibited from publicly developing drivers for KeyGrip. Within weeks Truevision was named as Microsoft’s key developer of ActiveMovie 2.0. CEO Lou Doctor told the press:

Windows NT with ActiveMovie 2.0 and Truevision's TARGA hardware will be a compelling combination for digital video professionals looking for superior price/performance...

Then Macromedia and Media 100 announced a major editing collaboration.Industry pundit Ron Lindeboom recalls:

Kathlyn and I were guests of Media 100 and Macromedia for a preview party to give a sneak peek at (KeyGrip) leading up to the 1996 Conference.

The KeyGrip application was to ship with Media 100's Vincent video engine:

KeyGrip will support professional features including instant playback of edits, real-time effects, real-time audio mixing and digital video effects (DVE).

The hardware to power KeyGrip was said to be:

A new extended architecture version of the industry-leading Vincent digital video engine will power the Media 100/KeyGrip offering. Extended Vincent architecture is a highly scalable digital video platform comprising signal processing circuitry, industry-standard bus and peripheral connectors, and an upgrade framework that provides for future, in-the-field enhancements of hardware and software capabilities.

In a press release, John Molinari explained the bold move.

Macromedia has worked closely with a broad range of video editors and cinematographers to ensure KeyGrip will meet the requirements of video and film professionals without compromising ease-of-use and accessibility for non-professionals. KeyGrip coupled with Vincent's unique digital video architecture will be ideal for anyone aspiring to deliver broadcast quality results from the desktop.

Macromedia's director of video products marketing, Tim Myers told the press:

Media 100's unique Vincent digital video engine is the ideal hardware platform for KeyGrip. Together, they deliver the industry's leading open systems solution for digital video authoring.

Anthony Dolph, Media 100’s director of marketing, concluded:

With QuickTime for Windows, Macromedia's KeyGrip software and Media 100's next generation Vincent digital video engine, PC developers can feel confident that they can save time and money while getting products into customer hands more quickly and efficiently.

When the Macromedia team returned from the Developer’s Conference they sat down for a brainstorming session to create a shipping name for KeyGrip.

Tim Myers recalls:

I had been through a few naming exercises both at Macromedia, and Adobe with Premiere. We got everyone on the team together and threw a bunch of names together, and then held a vote. There weren’t any marketing consultants involved, and when we landed on Final Cut we were all pretty happy with that as our choice.

John Molinari saw the agreement with Macromedia as his best shot at shipping an affordable editing system on the PC platform.

John Fierke had been with Data Translation as a senior engineer for 16 years. He was the obvious choice to lead a team, focused on making the Vincent hardware work with Apple’s QuickTime Windows API.

I moved from the Data Translation group to the Media 100 group to start the Bobcat project. I was the main technical contact with Apple and Macromedia.

A good-natured rivalry evolved between the PC and Mac engineers. Fierke recalls:

We initially thought that getting our hardware, firmware, and low level driver code ported from Mac to Windows would be a huge challenge, but this turned out to be surprisingly easy. The Media 100 Mac guys started out mocking the Windows engineers with comments like, “It’s so complicated, you’ll never get it working on Windows” but when we got it completed quite quickly, they took credit!

“Our design was so good” they said, “a monkey could have ported it to Windows”.

Fierke needn’t have worried that the Apple engineers might not be Windows savvy enough.

They knew, as much as anyone about the Windows pieces for QuickTime. The real challenge turned out to be dealing with a huge company 3000 miles away.

Final Cut’s Michael Wohl recalls:

Most of our engineers were Mac guys but the Windows version had priority for market share reasons and we worked to that. NT was pretty solid and it looked like with Apple's problems that it was definitely going to be NT.

Then Macromedia hosted a meeting with Sony Corporation of Japan. Wohl recalls:

We had spoken with Integraph about a complete turn key system and I even demonstrated a beta version on a turn key system for Toshiba at InterBee in Japan. Then we met with Sony and their video people because they wanted to buy the whole Final Cut operation outright.

Along with Tim Myers and others, I demonstrated the current iteration of the product and we got a very positive reaction, until the executives from Sony asked us what we expected to charge for a single license.

Macromedia had always maintained that it would ship Final Cut as a software package at around $5000 and let resellers bundle it with various hardware video cards for a total price of $8-10,000. Sony wanted to charge editors around $50,000 per licence.

The meeting ended after price was discussed. Sony showed no further interest in KeyGrip.