The Power of Hiding Complexity

During the last week I caught up with the (in my totally biased opinion) excellent take on FCP X and the Future of Editing by David Leitner in Filmmaker Magazine. He outlines the process of democratization that has happened since Final Cut Pro 1, and how that affects what is needed from a modern NLE. That article set me thinking about how hard it is to hide complexity in a simple interface and how that affects the learnability of the software.

The whole article is excellent but what really caught my eye was this excerpt:

What cracked the door to editing for “the rest of us” was the introduction of software-based Final Cut Pro 1.0 at NAB in 1999, coincident with the arrival of FireWire-enabled DV camcorders. All dismissed as amateur, naturally.

We’ve seen how that turned out. Digital democratization spread inexorably. It’s not hard to draw straight lines from FCP + DV to HDV and DVCPRO HD, to the rise of small camcorders, Internet streaming, cheap SD cards to record on, RED usurping film and cheap HDSLRs usurping RED. (With a regrettable drop in pay rates along the way.)

As a consequence, there are exponentially more people, professional and nonprofessional alike, of all ages, in all countries, now creating, editing and distributing digital movies. Everyone with a point-and-shoot or smartphone in their pocket is a potential HD source. (Today’s equivalent of a Kodak Brownie: “You push the button, we do the rest.” Insert irony here.)

Make it easy and it becomes more accessible to more people. In a conversation recently with someone I consider very smart, they mentioned they’d put together a full Final Cut Pro system, around the time of version 5, intending to do some editing work. He claims he never used it because it was too daunting. The worry was that if a setting was wrong at the start, he might not be able to output at the finish. Point being, it was so daunting that a very smart person was disheartened enough by the apparent complexity to avoid it completely. He’s also attempted to use iMovie and finds it too limiting (he’s not an editor by any means).

My first introduction to NLE was via Media 100: the very essence of a simple interface. I could teach everything about Media 100 – from turning on the Mac to outputting a project – in three days. Final Cut Pro 101 alone is three days (201 takes a full working week)! (Thanks to Robin Kurz for the correction.) But it was very easy to grow out of Media 100. Media 100’s metaphor was simplicity itself: an A/B video track with transition in between, with a superimpose/title track above. The flow of the video was clear from the arrows between the tracks. It was a simple interface and it appealed to me at the time for that reason: it was way more powerful than my cuts only suite that it obsoleted and the quality was high enough for finished work. It should be noted that I was totally self taught as an editor, and came into the industry via my own business. I was almost the exact customer that John Molinari had in mind for Media 100. The metaphor resonated.

Media 100 was easy to outgrow. It lacked the depth that we all reach for sooner or later. In my early days, After Effects actually provided the effects depth Media 100 lacked. (And of course, Boris built the entire BorisFX business on the assumption Media 100 users – and others subsequently – wanted more compositing power in their simple NLE interface.)

Fortunately about the time I was finding Media 100 limiting, along comes Final Cut Pro 1. But the metaphor was different. And different from what was originally conceived at Macromedia as a competitor to Premiere 5 (effectively). That tool was shown at NAB 1998 in a small booth off the show floor and I caught a demo of 0.9 alpha of this new Final Cut (and I have the T shirt to prove it!) That was very similar to the Premiere of the day which had A and B video tracks, with multiple superimposition tracks above. And indeed 0.9 of Final Cut (at Macromedia) had eight “After Effects style compositing tracks”.

But by the time Apple had reworked Final Cut to be the democratized editing tool for professionals (see my earlier post on how few units Apple expected to sell) Final Cut Pro was more Media Composer than Media 100. The metaphor made sense to those who had been editing with film or video tape in traditional edit bays. It assumed a certain level of knowledge about settings, timecode, etc, which was frequently a barrier to entry.

My partner Greg who has no video production background at all, understood Media 100, possibly because he did a lot of the publishing work on the Media 100 Editor’s Companion. But interestingly, he never really understood Final Cut Pro 1-7, despite doing a lot of the publishing work on the DV Companion and subsequent training products. In discussing this with him, it’s clear that the underlying metaphor for Media 100 made sense to someone without a background in video post production, whereas Final Cut Pro 1-7 did really require a background in post production to be comfortable.

In other words, Media 100 suited the producer, shooter, or “unschooled” editor; Final Cut Pro, despite its attainable price, really required the editor to be “schooled” in the traditional methodology. That metaphor wasn’t helpful to the hundreds of thousands of people editing, who had never done tape-to-tape editing (or the even fewer who had edited film on a Moviola). This is what lead to the impass I mentioned above, where Final Cut Pro was so daunting no start ever happened.

Of course, many more people purchased, and learnt to some degree, Final Cut Pro 1-7. The growth went way beyond Apple’s expected “25,000 seats in two years” and along the way trainers – such as myself – did good business helping people access the undoubted power of Final Cut Pro 1-7.

Faced with this change in landscape since the original design of Final Cut (Pro), when Apple came to rethink and redesign Final Cut Pro, it seems logical that they should not continue the metaphor that required a heap of industry background and specialized knowledge, or start over with an interface that allowed the newly democratized professional to quickly become efficient.

Now to be fair Adobe went to a lot of trouble in the CS 6 release of Premiere Pro to make the interface less cluttered by a lot of tools (power) even if they were infrequently used. Customizability in Media Composer similarly helps simplify the interface. However, in both cases the metaphors are still those of a 20-year-earlier industry.

My observation is that people – like Greg  who understood Media 100 – find no problem with Final Cut Pro X – it feels intuitive *unless* you have a bunch of specialized post production experience with traditional tools. I hear the same from friends who continue to train both versions of Final Cut Pro. Universally they say that it was daunting for a newbie to get up to speed with classic Final Cut Pro, but very easy to bring those same people up to speed with Final Cut Pro X.

What I find somewhat confusing, is that some of my smart, and experienced, friends, took much longer to adapt to Final Cut Pro X than I would have expected (or experienced myself).

Of all the efforts to simplify the editing interface, I think Final Cut Pro X does it best. At first look it’s superficial similarity to iMovie let people think it lacked power. But just one click or keystroke away, there’s a power feature. For example: if you were happy to use Keyword Collections as superficially similar to Bins in earlier versions of Final Cut Pro, you can by simply applying the keyword to a range. But if you want to create complex filters to dynamically pinpoint the exact clips you’re looking for, then you have that as well. (And if you want to ignore keywords altogether and just use the Favorite and Reject features that it shares with iMovie, that’s OK too.)

Final Cut Pro X is deceptively simple. You can work with the clips and do hard audio cuts with your video; or you can open out the audio track and do easy overlaps and cross fades. Or dig deeper and open up all the audio components and work at that level. (We could still use a role-based mixer, please!).

If you’re happy with the single viewer approach in Final Cut Pro X, then the Event Viewer never gets in your way. If you want it, it’s there in a flash. Preset speed changes are very easy, but just underneath you can adjust speed dynamically.

I could go on, but there are lots of examples. Clearly there has been thought put into how to keep the power, but retain as much simplicity as possible.

Another way they achieve apparent simplicity is by having what I call “sensible defaults”. Final Cut Pro X does by default what I (mostly) want it to do. Mostly when I trim clips in a timeline I want the timeline to close up. It annoyed me that the Roll tool was the default choice when I mostly wanted Ripple.

I think it is the combination of “intuitive” interface (as long as you don’t have a lot of knowledge already) and sensible defaults that makes Final Cut Pro X so fast, but I’m still trying to quantify why everyone who has become familiar with Final Cut Pro X says it is so much faster. (They all say it’s faster once they get used to the changed paradigms.)

So, Final Cut Pro X seems to be intuitive, but only to those who have not had a traditional media production background. As David Leitner says in the quote above:

As a consequence, there are exponentially more people, professional and nonprofessional alike, of all ages, in all countries, now creating, editing and distributing digital movies.

So the people without the knowledge of the historic paradigms make up the majority of the marketplace. Those people have no problem with Final Cut Pro X, but those familiar with the older paradigms have trouble unlearning?

Maybe unlearning has become as valuable a skill as learning. I drink wine to help with the unlearning, but perhaps Alving Toffler put it better:

The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn.