[Chris Kenny]

[Herb Sevush] “Since I never worked with Avid I can’t talk to all it’s cludgy grafting on. But I did work with a program that behaved like that, with an incredibly unintuitive interface that had user options so spread out through it’s incredible maze like menu trees that it was almost impossible to know where any feature was listed – it was called Final Cut Pro. “

Classic FCP started life as a Macromedia app, and honestly it has never quite been up to the standards of most of Apple’s software. The ‘Tools’ menu in particular is sort of a random grab bag, and in my experience some large fraction of FCP editors have never quite figured out audio routing.

But for just getting footage into the app and performing basic editing operations with the mouse, ‘classic’ FCP is worlds ahead of MC in terms of usability. I think this is a not-insignificant reason for its success, and I think FCP X is even better at ‘simple things should be simple’ (as long as you’re not defining ‘simple’ to mean ‘the same as they are in [other NLE]’), and that this will pay significant long-term dividends in terms of adoption.

More broadly, FCP X is simply of a different era from Media Composer not just in terms of UI and internal architecture — there’s also pricing (it’s ~80% cheaper) and packaging (it’s available for one-click App Store installation). The whole computing world is evolving. This industry exists within that world. And Media Composer — ‘heavy’, expensive software with a proliferation of windows and modal dialogs and a — is what things are evolving away from. Existing users, particularly those who are just interested in editing and don’t really follow the broader computing industry, don’t necessarily know or care about any of this. But it will absolutely impact what choices new users and facilities make over the coming years, even if the people making such choices aren’t explicitly thinking about it.

[Herb Sevush] “And then the powers that be re-wrote it from the ground up guided by all your UI rules with modern fonts and came up with a program that could do simple stuff simply but was totally unsuitable for a large segment of it’s customer base.”

But they’ve since turned it into something with the necessary capabilities to serve the overwhelming majority of their customer base, and without abandoning ‘simple things should be simple’.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Herb Sevush] “Or it can be seen as part of retro chic.”

FCP X’s main timecode display (which is clearly designed to look a bit like a sixteen segment display under glass, though if you pay attention you’ll notice it renders some numbers in ways an actual segmented display can’t) is ‘retro chic’.

Not using modern font rendering algorithms for text in bins and dialogs because you’ve never updated your software to do so is just ‘legacy’.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Herb Sevush] “#1 – What professional would give a sh*t.”

In some cases (possibly including this one), this kind of thing can be a visual indication of dated technical foundations.

[Herb Sevush] “The definition of an overly limited NLE would be one that a user could figure out without a manual or training. If it’s possible to do, then it’s not worth the trouble using.”

I firmly subscribe to the Alan Kay philosophy of software design — “Simple things should be simple, complex things should be possible.” Media Composer falls flat on its face with the former. Avid itself, and MC’s existing fans, are fooling themselves if they think that won’t hurt adoption. That’s bad news for a company that could really use some revenue growth, and it’s bad news for the product’s long-term market position.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Michael Hancock] “So post software should be built and judged entirely on how easy it is for someone with no experience with them to import footage, without any training? Seems silly.”

Seems like something I didn’t say.

I specifically argued that this was a problem for Media Composer with respect to attracting new users and expanding into additional markets. Not that post software should be judged entirely on the basis of walk-up usability.

[Michael Hancock] “Seems silly. How easy is it to get footage in Resolve for someone with no experience? Or Smoke? Or Speedgrade?”

I don’t use Speedgrade or Smoke, but I have found Resolve to be vastly less impenetrable, in the way I’m discussing here, than Media Composer. In fact, it’s hard to think of other examples of software as bad as Media Composer in this respect. 3D modeling apps, maybe. But they’re all like that. MC’s problem is that its major competition isn’t like that.

[Michael Hancock] “Do we completely disregard the reasons why an NLE prefers/requires a specific media structure simply because someone hasn’t learned why?”

Frankly, yes, we should disregard that when evaluating UI. “The internal structure of our app requires us to do this crazy, unintuitive thing” is not really an excuse, if there are other ways things could be structured that wouldn’t require that.

But this is exactly the sort of thing we expect when dealing with an app that has been around as long as Media Composer has. Really, think this through. When Media Composer was new, file-based workflow was effectively unheard of. Odds are your media files were being captured from tape directly into MC, so it worried about where to put them, and file-based importing was a special case, not a basic task, so handling it with a quirky workflow and a special tool was fine. Since then, file-based workflows, including workflows involving native camera media, have become common. So in addition to the system that was created for the tape-based era, Avid has grafted a whole separate system onto MC in the form of AMA, which references and works with media files in a completely different way.

I want the long-time Avid users to really think about this, as if for the first time. Media Composer has two completely different paradigms for working with media, depending on what format it’s in. Would you design an NLE like that if you were designing it from the ground up today? Not a chance. It’s purely an artifact of new capabilities being grafted onto a legacy architecture.

[Michael Hancock] “Are there things Avid could improve on, including interface tweaks? Absolutely. But to say it’s oudated is kind of silly. You’re not even a day-in day-out editor, Chris. Maybe the interface wasn’t designed for you?”

I’m not a day-in-day-out creative editor. I do quite a bit of online edit and conform work, including prepping projects in offline NLE environments. I also design and specify entire end-to-end workflows, from dailies, though offline, online and deliverables. I suspect a lot of creative editors who don’t see what’s wrong with MC may be focusing almost exclusively on a narrow set of core editing functions, and ignoring the fact that the whole structure around them has become a creaking relic.

The kinds of things we’re talking about here will not be fixed by ‘tweaks’. It’s worth noting that MC is the oldest of the major NLE environments, and the only one that, as far as I know, has never had a ground-up rewrite. The points I’m making here are a fairly straightforward, expected consequence of this fact.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Chris Harlan] “Drivel.”

Sit a new user down with Media Composer, FCP X, and Premiere. In each app, have a project/event already created and named appropriately. Give the user a folder full of video clips in whatever format is preferred by the NLE. So that would be DNxHD MXF files for Media Composer, ProRes MOV files for FCP X, and, well, Adobe doesn’t have an equvilant codec, but let’s just say ProRes MOV files there as well, which should work fine.

Ask the user to import the footage into the app.

With FCP X or Premiere, the user will see the appropriately named event/project and drag the footage in. This will work. Alternatively the user will look for some kind of ‘Import Footage’ command, and will find one and use it. This will work.

In Media Composer, you have to go create a /Avid MediaFiles/MXF folder, then create a numbered folder within it (numbers only, please!), then quit Media Composer, place your media in that folder, re-launch Media Composer, access the Media Tool, check all the correct boxes (which is a little confusing in its own right), and then drag the clips from the resulting window to your bin. This is completely undiscoverable, and, frankly, crazy. And it’s not an isolated example.

If you’re the sort of person who learns software by reading manuals and memorizing steps, maybe you really don’t see why this is any harder. If you’ve been using Media Composer for 15 years, maybe it hasn’t occurred to you to question this in a very long time. But if you’re a new user who learns software mostly by exploring it — which is how most people seem to prefer to learn software — Media Composer is very nearly impenetrable gibberish.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Chris Harlan] “Maize-like? Really? Doesn’t feel like a modern app? I’m sorry. I find these silly statements.”

I’m aware that lots of existing Avid fans (which I’d guess you are) don’t see any of this. It was rather the premise of my previous post.

I could point to some objective characteristics of the Media Composer UI that make it feel dated — its multi-windowed approach (there has been a strong shift in UI design to panes rather than separate windows), the style of its UI widgets and the fact that its toolbar icons aren’t antialiased, the fact that on the Mac it appears to still be using deprecated Carbon text rendering libraries (this is trivially easy to tell if you know what it looks like), its extensive use of modal dialogs, etc. And I can point to completely bizarre, non-standard UI, like the way it handles setting preferences or configuring keyboard shortcuts. I can also point to really odd workflow behavior, like the way it can only work with MXF files in numbered subdirectories in an Avid MediaFiles folder at the top level of a drive, and the fact that dragging an MXF file from the Finder into a bin does nothing useful — but oddly enough it will offer to do something useful if you drag in non-native media.

But really, it’s hard to make this case to someone who doesn’t see it, because it’s the sort of thing that you either see, or you don’t. I wear a fair number of hats in my present position, and I’ve worn even more over my career — I’ve done UI design and software development professionally, among other things. Media Composer simply does not work the way an app designed from the ground up by competent UI/UX people in the second decade of the 21st century would work. It works, frankly, like an app from a couple of decades ago that has had a huge number of features grafted onto it — because that’s exactly what it is. There are signs of this all over the place in, say, Photoshop as well, but Media Composer is by far the worst offender among apps I come into regular contact with in this industry.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Chris Potter] “Your R&D point is a great one. Without spending in this area it’s hard for the product portfolio to continue evolving. Without deep pockets it’s hard to invest in this area. Avid faces real constraints in this area. I believe that one of the risk factors that they mention in the annual report is along these lines:

“The market segments in which we operate are highly competitive, and our competitors may be able to draw upon a greater depth and breadth of resources than those that are available to us.”

It’s funny that they don’t mention people as a risk factor. I would be worried about the culture after so many restructuring plans and rounds of layoffs. I do think that if the new CEO decides to restructure in some form he will have to address people’s fears that it’s going to be death by 1000 cuts. “

It’s not just capital and employee talent. Avid’s ability to innovate is also constrained by their existing user base, which is surely the most conservative in the industry. Take Media Composer, for instance. I suspect pretty much everyone who hasn’t been using Media Composer for a decade or more (and some people who have) will admit that its interface is largely impenetrable to new users. There are lots of features hidden in there, but it doesn’t feel like a modern app, its UI is practically maze-like, and the ways it handles many things (like non-native media — a distinction that ideally shouldn’t even exist in a modern NLE) seem grafted on, because they were.

It seems unlikely Avid can grow the product into new markets and attract new users without addressing these issues. But addressing these issues looks like it would require a full-out rewrite of the technical foundations of the app, and a huge overhaul of the UI. And how do you do that without the existing, extremely conservative user base revolting?

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Herb Sevush] “It’s also not fast enough for some multicam work, even at 1080.”

Uncompressed, sure, but I don’t know that there are a ton of people trying to do that. It’s fast enough for over 20 streams of ProRes 422 HQ at 1080p/29.97.

Even in pro video, people who need storage faster than ~800-1000 MB/s are a pretty small niche.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Marcus Moore] “Thunderbolt in it’s current state is really just a much more robust Firewire. Until Intel can implement the much faster optical version that can support everything that PCIe can provide… the crossover point is moot.”

Thunderbolt supports PCIe ‘natively’, and while it’s true that it’s not as fast as, say, an 8x internal slot, it’s fast enough for pretty much everything except a) GPUs and b) real-time uncompressed RGB 4K video I/O.

It’s more than fast enough for 1080p and 2K video I/O, which by itself is a pretty big deal. It has really opened up pro video on machines without slots in a way that wasn’t previously true.

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.

[Chris Kenny]

[Aindreas Gallagher] “it looks, and sounds, just super clever. It’s actually, possibly, fundamentally dumb with regard to an editor’s cognitive and spatial appraisal and understanding of the various assets, but Apple themselves felt very, very clever when they decided to do this.”

Spacial organization — manual spacial organization, at least — largely falls apart as the number of items to be organized rises. This is probably why you’ve seen the spacial metaphors of the ‘Classic’ Mac OS finder fade away as well.

But perhaps Apple should look at ways to automatically create trees or clusters of clips laid out in 2D space based on metadata. Apple recently acquired a bunch of patents related to ‘axis-based user interface’ that could be relevant here.

(Although my suspicion is that their primary relevance is to some post-hierarchical file system scheme Apple is working on for iOS.)

—
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.