Production and Post Wars (or Why Red Should Buy Final Cut)

29 04 2010

Well, all right, I’m exaggerating there. I don’t really think that Red should buy FCP, and Production and Post aren’t exactly at war (though sometime you’d be forgiven if you thought that) but I want to make a point here.

Every year it seems that camera manufacturers create many “improved” codecs that answer their needs — increased quality with reduced file size. However, that goal is pretty much immaterial to post-production professionals. We don’t care if an image takes up a large file size. In fact, with the faster processors and cheaper storage costs (last I checked, a medium-ish quality 2Tb drive costs less than $300 on Amazon), we don’t much care what size the original file is. If it’s too big to use, we’ll just create a lower rez transcode in ProRes or DNxHD and edit with that. In fact, it’s more important to editors that it be easy to edit.

This means that Long GOP file formats, where most frames are not stored as full frames but as a smaller list of changes from the preceding frame, are horrible. They are exceedingly hard to edit with. Whatever speed gains we might conceivably get from working in a smaller file size are more than undermined by the extra work our NLEs need to do in order to display them.

[Note of ignorance. I haven’t yet had a chance to play with the parts of the new version (5.0) of Avid Media Composer which allegedly make a lie out of that last sentence. Pushing their Avid Media Access technology forward, and allowing the Media Composer to natively work in Quicktime, Red and various Long GOP formats, they promise to make editing much easier with these previously hated formats. This has proved to be true in my experience with the Sony EX-1 and EX-3 cameras, so this could be a great boon. And I’ll talk about that in a few paragraphs, so stay tuned.]

Let’s face it. Editors are never going to get camera manufacturers to stop looking for their version of “better” codecs. We’ve long since learned to live with it. But it does mean that, unless these manufacturers work ahead of time with the NLE manufacturers (the way Red did with Apple, for instance, before the initial release of the Red One) it’s going to take some time for our favorite NLEs to catch up with each new release of a camera codec.

It’s a war and the winner of that war is… well… no one. But the biggest loser is the filmmaker.

This is less of a visible problem on the bigger budget productions where the camera and editorial departments are made up of different people, each of whom have varying levels of tech support that go beyond typing “Long GOP won’t work” into a Google search bar. But as more and more of us are shooting with small crews, and taking it back into the editing room where we have to ingest and edit it (and output it) ourselves, this becomes more than an annoyance, it becomes an impediment to our livelihoods (you know who I’m talking about, you WEVA folks out there).

So, what’s the best solution to this war? Is hope for reconciliation only slightly less feasible than the Democrats and Republicans agreeing on anything in Washington today?

Well, yes it is. But there are some signs of hope.

I’ve already mentioned Avid’s AMA. What that does is create a set of open architecture hooks for camera manufacturers, so that they can more easily create a way for editors to edit natively in the Media Composer. It’s an attempt to make it easier to do what Red did with Final Cut before the Red One’s release.

In both cases, it’s the NLE manufacturers telling the camera manufacturers — “Hey, if you’re going to create your own camera codecs, you’ll have to create your own editing codecs.” Well, not exactly, but Apple and Avid are placing the onus on the camera manufacturers to dig themselves out of their self-constructed hole. And that makes sense, so long as your NLE is one that has enough of an audience to make it worth the camera folks’ attention. I might be wrong, but I doubt that Sony, Panasonic, Red and the HD-DSLR manufacturers are going to spend buckets of money writing plug-ins for Liquid or Vegas.

So, what are our other alternatives?

In the old days, every single camera manufacturer had to create cameras that worked with the industry standard 35mm film gauge. If they wanted to create a film that was a different width — such as, say, 38mm — they had to be able to manufacture the film, the lab processing equipment, the editing equipment and the projectors to accommodate that.

Needless to say, we never saw 38mm film. [We did see 16mm and 70mm film — which at half and double the normal size was easy for Kodak to manufacture film for. When it became clear how it opened up new markets, the camera, editing and distribution worlds came along for the ride (to greater or lesser degree).]

But what if a company could manufacture a camera and editing and distribution equipment (like Sony) and didn’t have their heads up their posteriors (like, uh.., like… oh never mind)? In a frighteningly anti-competitive way, they could then create a camera codec that worked fine in both capture and post production.

We haven’t yet seen that company, though if Red bought Final Cut from Apple (or MC from Avid, let’s say) it would certainly be a start in that direction. Please note, I have absolutely no inside information on anything that Red, Final Cut or Avid might be up to. For all I know, Apple is planning on buying Red, though that would shock me in ways that I can’t describe in public.

In the meantime, Red Cine X and AMA are two ways that post and production are attempting to bridge the gap. last time I looked, Avid wasn’t manufacturing cameras, which will make it more difficult to keep up with Red Cine X.

When Cisco bought Flip last year, I was hoping that we’d see some real synergy in the production and post areas. At the very least, I was hoping that we’d see some changes in the Flip that would enable them to interact with the web backbone much more easily. That hasn’t happened yet, and there’s no indication that it’s imminent.

But wouldn’t it be awesome if someone came up with a series of codecs that could take footage shot by a camera, make it easily editing ready and trivially distribution ready. By this, I mean more than projector-ready (something that I am hoping that Red Ray will pave a path for) but will make it easier to distribute files safely to theater owners, television networks, web distributors, mobile device partners, et al.

And, I’m hoping that these solutions are provided by multiple companies so we don’t have to be tied to one technology.

Whoever creates that chain will be the Dag Hammarskjöld of all things digital video, and their company will be its United Nations. Peace at last!



Techy Talk

12 04 2010

I’ve got a post percolating about the use of the iPad in education but it’s not really ready yet. In the meantime, I wanted to spend a post or two talking about some more tech-y things.

It’s so damned easy to get swallowed up by the technology in post production nowadays. About five years ago, no editor that I know was using the term “workflow” and now it seems that that is all we talk about. Codecs?  Why should I know about them?  Well, honestly, it’s because that knowledge helps us to do our job better.  When I was a wee assistant editor, I made it my business to learn how the film optical houses did their job, as well as the labs.  I learned about white core mattes and black core mattes, so I could talk more intelligently about them when I was conveying our requests.

Now, take that and multiply by a thousand. I’ve talked before about how we need to know VFX, sound design, color correction and much much more in our editing rooms. Sometimes it seems overwhelming. Luckily, there are tools out there to help us do our jobs better.

Color correction is one thing that continually stumps me.  My wife, in fact, thinks that I’m color blind; she often stops me as I am on my way out the door in the morning with a “You’re wearing those together?”.

So, when Christian Förster, over at Avid Screencasts podcast, recently posted three separate vidcasts about color correction on the Media Composer I devoured them.  I waited until all three were released so I could watch them at one sitting and it was well worth the while.  You can go to his website, Avid Screencasts, to see them (as well as a number of other valuable episodes) or go directly to any of the three episodes here:

Color Correction Basics I – Laying the Groundwork

Color Correction Basics II – Manipulating Contrast

Color Correction Basics III – Manipulating Color Balance with Curves

Hey, Christian, you should put these three casts together into one, add some deeper discussion (primary vs. secondary for instance) and then sell them.  They’re that good. I’m going to put the three together for some of my classes.



Rodney Charter, HD-DSLRs and Prep Time

3 04 2010

Rodney Charters shooting using a Canon 5D

The one thing that you can predict with students is that, if there is a cheaper way of getting to an end product (and that way involves a cool new toy) then they will be all over it. At USC, a recent trend has been away from the Red camera (which was all the rage for the last 18 months) and moving towards HD-DSLR cameras, still cameras which have been tricked out to shoot video. Because they have the large image sensor of still cameras, the HD quality they can deliver is amazing, though there are still issues with production and (particularly) post production.

I have no doubt that companies like Adobe, Apple and Avid will eventually work out the post production issues, so it’s important that we all get familiar with the issues involved in HD-DSLR production. But there are much deeper issues here than, simply, the technology. Issues of aesthetics and storytelling aside, shooting with a DSLR camera isn’t the same as either a no-budget or a high-budget shoot and it pays to think about why and how.

Here is an article by Rodney Charters (best known for his work on the show 24 and followable on Twitter with the handle @rodneykiwi) about the shooting of an Indian gangster (post-Sopranos) short. You can get to the article on digitalphotopro.com.

What is often left behind on these sorts of low budget shorts, and the article gets into, is how misleading the camera is, in terms of prep. And that is exacerbated on an DSLR shoot. Because it looks like the point and shoot digital still cameras that we’re used to taking out and capturing family picnics with, there is a tendency among many new filmmakers to treat their own projects a bit too informally. Crews aren’t bothering to do camera tests before shooting, and very little concern is given to issues like how lenses affect focus, and how handheld shots on a small camera are differently executed than on a larger one. This leads to beautifully detailed HD images which are slightly out of focus or too bouncy to use (remind me someday to talk about the advantages and disadvantages of the Stabilize effect in our NLEs).

But there’s a mention in the article of some of the prep work that Charters did for this short, directed by Snehal Patel,

Working handheld or on a tripod, getting proper focus is a major obstacle for many filmmakers working with the 5D Mark II’s full-frame image sensor, which is closer to the size of 65mm motion-picture film. Because Patel and Charters were predominantly using Panavision zoom and prime lenses rather than DSLR lenses, it made the job of focus pulling a little easier. With cinema lenses, the extra-large lens barrel spins nearly 360 degrees, which makes it possible for a focus puller to hit critical marks within inches rather than feet. Similar to a film shoot, they also employed a first and second AC on each camera like a professional movie production.

Earlier in the piece, it mentions that Charters had one day of testing for a two day shoot. In other words, 1/3 of their production time was spent in testing. I’m not saying that Charters didn’t already know that he was going to using those Panasonic zoom and prime lenses, but I don’t doubt that what he learned in that one day helped the project move more smoothly, build the language with the director, and help the project to look better.

[As a side note, I should also mention that Charters, being the Hollywood DP that he is, had Panasonic build him a special PL lens mount so he could use a 10:1 zoom. Just try that if you’re Mr. or Ms. Indie Cinematographer.]

The piece also mentions that they had a DIT on set, and that they ended up backing up their footage to three locations. That’s professional industry practice on file-based cameras, but there are all too many occasions where crews shooting with HD-DSLR cameras forget that they still need to think professionally. They aren’t operating on a home movie shoot. Unless you can afford to lose your shoot like a home picnic, then you can’t treat it like one.