Seen today, the EditDroid It’s not even archaic, it oozes power and potential, like a Spitfire, but what’s common today was out of this world in 1984.
In 1984 the art of film editing was basically the same as in Louis Lumière’s time. The main equipment was a razor, glue and tape. Films were inspected manually, under a lamp, until the editor found the scene he wanted. He would then cut the film with a blade, and paste the scene into the previously edited scene.
Imagine a jigsaw puzzle where you have to manually cut out the pieces before starting to assemble. Rolls and reels of film, each with multiple versions of the same scene, handwritten directions from the director, tapes of scenes hanging from clotheslines, and the editing room floor littered with (literally) cut scenes.
In 1924 this work became a little easier with the invention of the moviola, an equipment that made this work a little easier. Now the film was inserted into a kind of mini-projector, with a screen that was also small. The editor used pedals to advance or rewind the film to the desired part. When he found the ideal sequence, he marked it directly on the film, with a crayon.
Now imagine that most directors shoot a lot more than they use in the final version. Apart from scenes with variations, we have scenes in which one actor was very good but another was more or less, a good editor will cut the shots of each one and form a new dialogue.
A 30-second commercial can require two hours of footage, including variations, retouching and experimentation. Some directors are famous for shooting versions and versions of the same scene. Charles Chaplin commonly passed 300 versions. In this video here Stanley Kubrick directs one of the 127 shots of the freezer scene in The illuminated.
Blockbusters, which cost a fortune, often shoot dozens and dozens of variations, experimenting with different dialogue, placements, and even switching characters. One Avengers: Endgame da vida ends its footage with a ratio of footage of 400:1, that is: for every 400 minutes filmed, one will appear in the final edit.
We’re talking miles and miles of movies. This was complicated even for modern 1980s moviolas, and video wasn’t much different.
The standard video editing setup consisted of a cutting/effects table, two or more players and a recorder. Editing was equally linear, you had to advance the tapes to the point where the desired scene began. You marked the beginning, then the end, hit play and the recorder would record that piece, maybe with some video effect, or input a second scene, coming from the other player.
Some editors love Moviola, find it pure magic to physically edit a film, others crave faster and more practical methods. One of these was George Lucas, who suffered horrors from Star Wars which, by the way, was saved in the edit, thanks to the efforts of Paul Hirsch, Marcia Lucas and Richard Chew (Bacca) – Sorry, I couldn’t resist.
Lucas dreamed of an editing system where the entire film would be on a computer, the editor could select excerpts without having to go forward or backward in physical films.
Of course, in 1984 there was no DaVinci Resolve, Blender, Inferno, Sony Vegas or Adobe Premiere, if you were to travel back in time and try to install Final Cut on a newly released Macintosh it would explode like the proverbial elephant joke.
Storage was another issue; in 1983 a high-end IBM-XT came with a 10 MB hard disk, hours and hours of digitized video can’t exactly fit into 10 MB. Anyway, a regular PC wouldn’t do it, luckily Droid Works, a Lucasfilm subsidiary that won a blank check to design and build the system, thought aloud.
They chose as the basis for EditDroid a Sun 1 workstation, released in 1982, with a Motorola 68000 CPU, 1020 x 800 monochrome resolution, 640×480 in 256 colors, up to 4 HDDs of 84 MB and 256 KB of RAM, expandable up to 2 MB . Okay, I know, your watch has more power than that, but those were different times.
Obviously EditDroid couldn’t read films on film, let alone process them.
A frame in 2K, the minimum acceptable resolution for a very cheap digital film, has a resolution of 2560 x 1440 pixels, totaling 3,68,400. Each such pixel needs color. RGB, 8 bits each, 24. That is, a single frame in 2K quality occupies 8.3 MB, 4 times the total system memory capacity.
The solution? EditDroid wouldn’t do direct editing, it would use a low resolution version, a trick used to this day when dealing with 8K or even 4K videos, depending on the capabilities of your powerhouse.
And where would these lower resolution movies come from? Simple: to guarantee agility, all the RAW footage of the film to be edited would be digitized and converted into videolaser, a format that does not degrade and can be replayed infinite times, and on top of that it was very fast, with control and access frame by frame, granting the editor full control over the sequence.
Of course, the process was still extremely slow. It took days and days, maybe weeks to digitize the films and press the videolasers, all the data had to be entered (hey!) into the EditDroid database, but after that the editor had everything that had been filmed at his fingertips, automatically accessing entire walls of racks with laserdisc players.
EditDroid came in the configuration with three monitors: one with the graphical interface, another for preview and a larger one to display the “rendered” version of the edited sequence. A special control featured several dedicated keys, a trackball it is a jog shuttle to fast-forward or rewind the film, but unlike systems that used magnetic tape, EditDroid was a non-linear editing system, you could go to any point in any sequence, right away.
It was also possible to apply chroma-key effects, cuts, scene changes, fade-in, fade-out, dissolves, and many others. This was especially important, as in traditional editing any scene change other than a simple cut would have to be specified, the parts sent to the photo lab and only days later the editor would receive the film with the transition.
EditDroid, besides being versatile, was extremely fast. In one demo, an operator using EditDroid competed with a publisher using traditional methods, the goal being to edit a sequence with four or five scenes. EditDroid finished in five minutes, publisher took 20.
As the images on the videolaser were broadcast quality and EditDroid used professional recorders, he was able to edit a TV show and deliver a tape ready to air, as well as films, then it was more complicated.
As the editing was done with low resolution proxies, the output was a list of cuts. When the final version of the movie was rendered, shown and approved, EditDroid would print an EDL – Edit Decision List, a list that indicates which part of, which video will be used, which moment to start, which moment to end, cuts, which audio in the scene, which roll of film the clip is on, everything.
This list is passed on to an editor, who will then have to assemble the film in a traditional but much more efficient way.
EditDroid was admittedly efficient, futuristic, revolutionary, even too much. Which explains the cold reception.
Editors, like everyone else, don’t like to wake up to find that everything they knew is useless, and EditDroid was a completely new tool. Movie people, used to working in a small room with the smell of film, had no interest in fiddling with computers, something far less common in 1984 than you might think.
The equipment also had another problem: The price. An EditDroid cost in current values $428,000. A moviola with cutting table and all the accessories, would go out today for US$ 31 thousand. Not everyone could afford nearly half a million dollars in investment.
By 1986 Droid Works had only sold 15 EditDroids. Lucasfilm had not yet produced any feature films with EditDroid, and the equipment was restricted to television, where it was used in series such as The Young Indiana Jonesbut the lime shovel was technological evolution.
Hard Disks with greater capacity and lower prices began to appear, editing systems such as AVIS abandoned tapes and did not even consider videolasers, now the films (or tapes) were digitized and transferred to HDs. It was all much faster and more agile, and the little interest in EditDroid disappeared like tears in the rain.
George Lucas realized he would never get a return on the $115 million (in 2022 values) invested in the project, and closed Droid Works in 1987. Avis purchased the EditDroid software in 1993.
EditDroid started the non-linear editing revolution, using the most technology of its time, but it was too advanced, too complicated, and too radical to be accepted. In the end, only 24 units were made.
The complete story of EditDroid can be seen in the great documentary EditDroid Rise and Fall.