Скачать книгу

Lab and stacked them three high. Bruce Leak adds:

       It certainly showed how complicated and how hard this was. They couldn’t even always render an entire frame, they had to render pieces and assemble them on a file server….[then] they had to load 100 frames at a time onto a digital video storage machine and output that to videotape.

      Apple Fellow Al Alcorn told the Computer History Museum's Henry Lowood:

      "...the fun was we were going to do the rendering as a desk accessory in the background at night on all these computers and the Internet, the network had just come out in engineering. Under Steve Jobs, there was no network. Now we actually had an Ethernet network emerging in the lab, and so everybody could put this program called NetSlave that would do the rendering at night. But [after] all this work, [we] realized that, wait a second, the Mac operating system is a terrible platform for video, for media, just terrible, terrible."

      Mark Lentezner and John Worthington designed, created and added audio but during this phase there were problems synchronizing sound and video. A small team called 'The Time Lords' spun off to solve the issue. Al Alcorn continues:

       And so, a group was formed called "The Time Lords" in engineering. It was basically an interest group that would meet every week and talk about "what do we do to fix the operating system of the Mac to allow for media to get in there?"

       That evolved into, that infected product development, and that became QuickTime.

      175 BEDFORD STREET - 1988

      Warner was determined to make Avid Technology a legitimate business concern and on the 13th of January, 1988 Warner leased 3000 square feet at 175 Bedford Street in Burlington, Massachusetts (above). Warner and Jeff Bedell walked into the machine shop that was laid out in a typical workshop fashion.

      Warner took the foreman's office that looked down to the factory floor and loading dock below. Bedell set up two loan Apollo workstations on the oily floor underneath old wooden beams.

       When Jeff used to type on the keyboard you could hear it echoing throughout the entire room - clack clack clack.

      Bedell had five months to make a working prototype but almost immediately discovered a problem

       The Apollo machines had the horsepower but they just couldn't match DOS or the Mac for simplicity and you needed a simple operating system to give you the room to get the video throughput happening. You needed a minimum throughput of 100kb per second for decent recognizable images and the Apollo was problematic.

       It had been built quite rigorously, but there was layer upon layer upon layer of code and when you got to the point of adding new code on the top, the speed was gone.

      Bill Warner explains what Bedell was trying to achieve with the Apollo workstations.

       When Jeff and I started, the immediate plan was to achieve a system that could replicate what a normal CMX online bay could achieve in the first release! I figured we could write and replicate editing and 2D effects like A.D.O. and dissolves and wipes like the Grass Valley switchers and titling like Dubner’s CBG, all in one box. I thought that surely there isn’t that much to re-creating a process that already exists. You could see the whole online video process as it existed then in any CMX suite.

       In other words we didn’t have to imagine what was needed by editors because we could already see the video and audio processes that were required. The difference of course was that, we wanted to take an analog process and make it digital.

      Bedell began work on an alternative user interface based on Pete Fasciano Oz notes. Naturally the VizWiz cofounder was one of their first visitors.

       I drove out to the Burlington factory and Bill’s showed me how they had progressed the initial HyperCard model that was called ‘V Edit’. It had a strong graphical interface that reflected the mechanics and iconography of a film editing suite

      Warner adds:

       I thought that once we had a working editor then we would learn about the film industry and move onto conquering feature films, straight away. One of my first motivations was to have a movie credit for Avid. I imagined seeing our company logo at the end of a major Hollywood feature. I was pretty naive looking back.

      Jeff Bedell recalls:

       I created an updated version of the filmstrip approach UI from Rick Olha's early iteration with Bill. It was essentially a digital version of dragging a filmstrip across a light box, the approach that Quantel had used with some of their product. It showed all of the frames of a scene together visually, but Pete shook his head. He said ‘this won’t work for video editors’. It doesn’t show you the temporal nature of editing, the flow of editing. It’s fine for still frames and rotoscoping but not for video editing.

      Fasciano continues:

       Bill was very interested in what I had written in the Oz document and Jeff had also built a basic ‘straw man’ version of my Oz brief. At this point I was just an invited guest giving them a number of new thoughts as an extension of what they were already doing. I knew it was critical for linear video editors to use time code as their method of communicating with their material.

       Of course the Avid would need to be frame accurate but it also needed a way to have a transition editing interface where one could trim a frame or add a frame with a plus or minus keystroke. So it was about fine tuning an edit.

      Bedell, with Eric Peters working part-time, refined the two UIs running on the loan Apollo computers for the up-coming NAB in Las Vegas. Jeff Bedell recalls:

       I had two versions of the editing model running on one machine at any given time. Why? You may ask. Because Bill was still out trying to get funding for the company and he was likely to just arrive back at the office without notice and say, "OK Jeff show us the latest Avid build! Show us where you are up to and how cool the system is”.

       I would have a relatively stable branch of base code running that I could demo with either of the UIs, the filmstrip or the Oz and then I had the current build of the model that I was experimenting with, the one that was a later software build but was prone to crash. When the demo finished I would switch back to finessing the build number 2 to make it stable and reliable.

       Then I would hand it off to Eric or one of the part-timers we had working at Burlington. Then when it was stable I would dump Bill’s demo version and start afresh on a new ‘scary branch’! And when I say scary, most of the advances that were on that branch weren't new things or new features but were more things that could last a long time, getting objects in place.

       All the time, thinking about our deadline, all the time doing demo’s for Bill and Curt, all the while handing it off to be checked.

      Fasciano adds:

       I wouldn’t say their other version was a film strip as much as a bunch of pictures up on the screen that looked like a piece of film. Perhaps his marketing background told him that making the UI look impressive may be better than a more sparse UI that works more efficiently. Either way what they had wasn’t going to work in the long term because this timeline had too many pictures in it and I wanted them to succeed.

       I told Bill and Jeff, “You can’t zoom in and out on a timeline with that many images because there just isn’t enough screen real estate to do that and besides it takes a lot of computer power and coding to make it happen.

      

Скачать книгу