Digital in the Desert

Day to Day on Episode II

Last fall, George Lucas made a bit of history when he left the 35mmequipment behind and took Sony/Panavision 24p cameras into theAustralian desert to shoot Star Wars: Episode II. Lucasfilm hadalready stirred up the digital pot in 1999, by premiering EpisodeI on four different digital projectors in New York and LosAngeles.

In this interview, Millimeter contributing editor Pete Putmantalks to Lucasfilm producer Rick McCallum about the Australian HDshoot, digital dailies, digital production issues, and the currentstate of digital cinema.

Pete Putman: You shot for 60 days--was that what you hadbudgeted originally?
Rick McCallum: Oh, yeah. We never shoot more than 60 days on ouroriginal shoot, then we cut for five or six months and then go back andshoot for a week or two weeks, then we usually go back a second time inOctober or November, and then go back for the last time inFebruary/March, just before release.
On Episode I, we did probably roughly two to two and a halfweeks of reshooting throughout those three different periods, andthat's roughly what we'll do on this. We did two weeks in London abouta month ago; we'll probably do anywhere from two to three days inOctober and November and then we'll probably do a day or two inFebruary/March.

PP: When you went out to shoot digitally on location, I'msure you anticipated a lot of things that could happen. What impactsdid shooting digitally have on your entire production process?
RM: There was nothing negative in shooting on digital and a lotof the positive issues I can be very specific about. First of all, tobe able to reload the camera only twice in a day--in other words, youput a tape in the morning before you start shooting, and at lunchtime,you take that tape out and put another tape in. I know that soundsridiculous, but you've got to remember, the way in which we shoot-- weshoot 60 days--it is imperative for George to have material. He doescoverage, but he also shoots very quickly. We have rehearsals.Everybody knows exactly what their job is each day, and we rarely everwalk away from a day's shooting without accomplishing at least 36setups a day. In a 12-hour day that means basically every 20 to 30minutes we're moving a camera and we're on a new setup. If we'reshooting with two 35mm cameras, and we reload 18, 19, 20 times a day ormore, it still takes three to five minutes to reload, reset, recapturethe moment again. We don't have that problem anymore.
If we're in the middle of a take and George wants to keep on goingbecause the actors are struggling with marks or struggling withemotional issues, we just keep rolling.
The ability to have a tape in the Avid 60 minutes later is justextraordinary, especially if you're shooting in five differentcountries, like we do. I can't send rushes back from Tunisia, get themto London, develop the negative, then have the negative sent to atransfer facility, then have those tapes sent back to me in Tunisiawithout having that done in three to four days at the earliest. Now,it's in our editorial system instantly.

PP: You had no problems with the camera?
RM: None whatsoever, and that's the other thing I want to gointo. That camera housing has been around for five-plus years, and it'sbeen used by every major network and every single known of situation:rain, sleet, doesn't matter what. We did not have a single cameraproblem, and I have never been on a film where that's ever happened. Infact, I have never gone into the desert without an engineer.
On Episode I, when we wrapped at night, our camera engineer wentthrough all of our cameras all night, and he'd have them all cleanedand ready for us the next morning. We did not have him on EpisodeII. We did have plenty of cameras as backup, but remember we weregoing through the first evolutionary stage of making a digitalmovie.
There were issues about negative insurance because there's nonegative. So to make it more palpable to our insurance company, wecreated a clone HD master simultaneously while we were shooting with asecond HDCAM machine. The one additional person we had was a videoengineer, but we had one less loader. It's a wonderful feeling, whenyou're making a movie, to have two versions of your film side by sideor in a different room.
One of the great things is we’ve shot 290 hours worth of [HDCAMfootage], which is probably equivalent to somewhere between 1.2 and 1.5million feet of film.

PP: I was going to ask you, as far as the actual camera crew... ?
RM: Exactly the same. In fact, one of the great things thatPanavision has done is that it has created the lens, the matte box,follow focus, everything. Other than to someone who actually knowslenses very well, it is virtually indistinguishable from a film lens.So we had no prime lenses. Well , we actually had one 5mm prime, awide-angle lens. The rest was shot on a 4:1 zoom lens and an 11:1 zoomlens.
The lenses are absolutely extraordinary. Anybody who has ever shot inanamorphic knows it is such a pain in the ass to actually get a greatset of lenses put together. To be able to shoot in high-def with theselenses that are so fast--they're 1.6 aperture for both lenses,including the 11:1 zoom …

PP: That's very shallow depth of focus ...
RM: You have all the depth that you want. And you have absolutecontrol. You know, we used a 1/4 pro mist on almost everything to beable to cut down the depth of field, but you can increase that to 1/2or 3/4, whatever you want. But if you want the depth of field, you'vegot it.

PP: These were anamorphic lenses?
RM: They're not anamorphic lenses. They're just basicallystraight because you pull the extraction in the frame wherever you wantit, which is another great thing. It's like shooting on VistaVision,you know, where I can pull an anamorphic frame anywhere within theVistaVision format.

PP: VistaVision was 8 sprockets horizontally like a 35mmslide.
RM: Exactly, but you're not defined by 35mm because you can pullyour extraction anywhere you want to. In other words, if you realize,"Oh my God--I don't have enough headroom on this actor," well, you canjust move that framing up further, or lower, or whatever you wantto.

PP: You shot to protect the full 16x9 [HDCAM] aspect ratio asfar as cables and everything else, but you were going to take aCinemaScope slice out of that, so you had some headroom top andbottom.
RM: Exactly. You have this huge frame and within that frame youcan pick your anamorphic--it's right down the middle, and that's how weframe to it, but there are many instances where George will say, "I'dlove to be able to pan up from that and come down," where we didn't panoriginally. Well, we can do it now digitally, within that frame.

PP: Did you have a CinemaScope graticule [in the viewfinder]that you were trying to stay within?
RM: On our monitor, within the eyepiece, and also in theengineering room. That's where the importance of having a person whocan not only replay, but also deal with the engineering side ofhigh-def, is going to be a great new learning step. Most video-systemsguys are very technical anyway, and it's not going to take long forthem to be able to accommodate all the additional stuff that needs tobe done.

PP: What was your shooting ratio?
RM: There is no ratio. Those issues go straight out the windowbecause you're paying about $68 for a 50-minute tape, and just to giveyou an idea, we basically spent about $18,000 to $19,000 for our"negative." Remember, there's no transfer cost. Imagine: Because we'rein the digital realm, every single shot is digital, we're in the Avidinstantly, we have no negative processing, we have no daily processing,and we have no transfer costs.
When you're shooting 1.5 million feet of film, you're looking at 2million bucks to go through that little routine. And that includes theshipping, the transportation, of getting the negative picked up at thelab, being sent to a transfer facility, the tapes involved with that,the transfer itself, the shipping back, and then getting the negativeback to the lab ... all that's a big, big deal. People forget about theunbelievable pain that it takes to actually shoot on film and get animage into an Avid.
On a more aesthetic level, when you're a director, to be able to watchyour image on two plasma screens, two 42-inch plasma screens, one foreach camera …

PP: Which models did you use?
RM: We used the Sony consumer panel, and it's great. We had noproblems, and like I said, we shot in five different countries, in theworst desert conditions, 125 to 130 degrees. We shot in a rain set forover a week and a half, just solid rain, with cranes, cameras,everything. We shot and moved the gear ... four countries in threeweeks--you know, massive moves--to Spain, to Italy, Asia, England. FromAustralia to Europe.

PP: When you were screening [dailies], I understand you had aChristie X4 DLP projector shipped out to Australia.
RM: It was right next to the editorial room, and [we] wentstraight from the Avid straight into the Christie … absolutelywonderful.

PP: Were you able to calibrate that projector to the sameHDCAM lookup and gamma tables you were using for everything else?
RM: Not at that particular time, but the image [quality] was sobeyond anything we've ever been able to see before. At the time, thewhole concept of our gamma tables was really interesting because you'retrying to set up, on a day-to-day basis, electronic wedges from whatthe image is going to look like when it's scanned onto film. And webasically said, “We know we can do this.”
We knew it when we did all of our tests, and we didn't transferanything to film until August when we left for Europe. And wetransferred about 4 1/2 minutes, 10-second shots from every scene, andwe were blown away. It was basically a one-light print--we're talkingabout no experience for anybody. But we were totally blown away when wegot back, when we were able to see it at ILM, projected on the TexasInstruments DLP Cinema system.
I'm not interested in which format is better anymore. The only thingthat drives us is that digital is indistinguishable from film. One ofthe reasons is that it doesn't matter what you shoot on what format.When you think about the film gauges, and the formats that have beenaround--whether it's early wide- or large-screen format or Cinerama,CinemaScope, Panavision, VistaVision, Superscope, Todd-AO, evenDimension 150--there have probably been 60 or 70 film formats since thebeginning of time.
If you look at the history of widescreen and 3D formats or film gaugesor different soundtracks, it's always been about quality. And it'salways been a group of individuals who have driven that within thecontext of the industry. One of the most tragic things about theindustry as a whole is that we have so little R&D from studios. Itdoes come mostly from companies like Panavision, Technicolor, Deluxe,Dolby, THX, wherever these companies are. They're usually small frytrying desperately to do one thing: have an audience actually see thefilm in the way in which the filmmaker made it.
And what's so brilliant about what we think we've got now, is this: Ifwe're really lucky we'll have anywhere between 250 and, if there's amiracle, 500 [digital cinema] screens out by 2002. A dream--certainlyfor me and I know for George--would be to have 5,000 screens by2005.
We will be able to scan out a 2,000-foot negative by next year, with nosplices. We go straight from a data file to an intermediate negative.It's a show print. We will be able to do 5,000 show prints. They willall be show prints. No one's ever been able to do that.
You know, if you're lucky, you can get five prints off the originalnegative. They're shown in New York and L.A., maybe Chicago forcritics, but that's it. They're not released; an audience doesn't getto see them. And we're going to have 5,000 show prints out there.

PP: These would be [struck] directly from the digitalfiles?
RM: Basically, our whole film is on a digital file, and thatwill scan out to an IN. We forget that we don't go through the originalnegative process, to an answer print, to an IP. We go straight from thedata file to an IN and from an IN to release print.

PP: You're only down two generations.
RM: Basically, yes. And it depends on how you view an original35mm negative. Is there a generation loss going from? We think thereis. It's complicated in the sense that each time we scan our data fileout to film--we have not only our lookup tables, we basically time thepicture as we go--we also have this "new thing" that we have to dealwith, a digital intermediate, which will allow us to hopefully get onthese 250 screens.
We don't know who's going to have the screens, there are so manysystems, so do we have to do digital intermediates for each differentprojector system? We don't know that yet. The release date forEpisode II is next summer; we don't know what day yet.

PP: A lot will probably change by then.
RM: Absolutely. But we'll be prepared, whateverdirection’s out there. I don't even think there will be an issueabout who's a winner or not because I don't think there's going to be ade facto standard by then. But whoever's out there withhardware, we'll be ready to match them with that digitalintermediate.

PP: So if you come out with intermediate strike prints fromthat, you're already ahead of the game by two generations.
RM: Absolutely. Two generations. Again, I don't have thishang﷓up about film--I love film. But what I passionately careabout and what George has been obsessed about since Star Wars,is quality. He actually does go to neighborhood cinemas three or fourtimes a week. And, you know, it's tragic. You can go to amultiplex--right now is a perfect time to do this, in the summer. Gosee Pearl Harbor. Even if it's run on a platter system,projected from one theater, even though it's the same print, onetheater to another theater to another theater, you will not believe thedifference. You'll get whiplash.
What we're trying to do is get rid of that inconsistency. That's why,at the end of the day, the issues are not about anything to dotechnically, other than that audience actually seeing a film in the wayin which we made it.
The real issue is not so much the acquisition of the images, whetherit's film or digital, but how does it get to an audience? How do theyactually get to see a movie that has the inherent qualities that all ofus spend millions of dollars and thousands of hours to try andacquire?

PP: There have been some comments made that digital HD[acquisition] formats don't have the dynamic range of motion-picturefilms. With film, if you are underexposed or overexposed in a shot,quite often you can figure out a way to save it. But with HD formats,you may not have the same dynamic range, and if you're under- oroverexposed you may lose some data. Did you find that to be thecase?
RM: It also is an issue of dealing with mediocrity. You know, ifyou have an untalented cameraman, it doesn't matter which format he'sshooting in. There is only a finite amount of talent. We have beenshooting in film for the last 65 years, and just tell me this: Why isthere so much junk out there? We're the only country in the world thatshoots only on 35mm for TV, and we have some of the worst lookingtelevision in the world. That has nothing to do about format. That hasto do with talent.

PP: What you're saying is: If you've got a goodcinematographer, and you're lit correctly, and you're taking readingsfrom shadows to full lights and making sure you know what the dynamicrange of the Sony HD cam is and always staying within that range, thisisn't an issue when you do compositing.
RM: There's another issue to shooting digitally. You actuallyknow on the set what you have. You don't go through the alchemy, whichis almost like fermenting wine. You don't know what you have with 35mmfilm until the next day. You think you know, but you don't actuallyhave it.
I guarantee you: Any cameraman that says that you actually do isabsolutely lying. If you shoot 1,000 feet of film, you just shoot aface, under completely controlled lighting circumstances, and you justroll the camera, without moving the camera whatsoever--you just roll,and shoot for 10 minutes, 1,000 foot roll--and you take that roll offilm, that negative, and you deliver only 100 feet a day over 10 daysto a lab--I don't care which lab it is--and then you splice those100-feet segments together … I guarantee your neck will be sorewhen you're watching them in the theater because you'll be all over theplace.
And that's part of the wonderfulness about film, also, because it isalchemy, it's chemistry, and every day it changes. And it changesdependent on how old the bath solution is, what the temperature of thebath solution is, how many times they change it, what the outsidetemperature is, what the outside humidity is--those are all different.It's like a wine coming from one part of Bordeaux, and a mile down theroad, on the same slope, is a totally different wine. You never knowthat from season to season.
But one thing you do know with digital is exactly what you have, on theday when you're shooting it, right then and there. You have that image,and that does not change if you're going to project it on a digitalscreen through a digital projector. One of the other great things is wealso have complete control. We did not have one out-of-focus shot onEpisode II, and we had a hundred on Episode I because wenever printed dailies. We couldn't print dailies. When you're going to30 different cities, you can't afford to stay in Italy for two dayswhile your film is processed to check it. We don't have thatluxury.

PP: It doesn't look like there's any down side to shootingdigitally--it's all upside.
RM: Well, for us, it was. The other thing we had was software,which allowed [cameraman] David [Tattersall] to be able to--while he'swatching the scene being filmed in front of him, he's looking at amonitor--click on any image anywhere while he's recording, and theimage is saved as a file. These files are used for matching, and wealso download those to Industrial Light + Magic so they can actuallysee film and the lookup tables with each image.
When and if we needed film--we didn't in this particular case--theywould be able to strike a print instantly and have an immediate visualreference of what the film has to look like. Like a clip test. It'sfantastic. You save it, so if you're shooting on, let's say, a largeset, and then you have to move to another set that's redressed andthere's certain things you want to do as pickups, to be able to havethat file available instantaneously is just wonderful.

PP: When you get ready to shoot Episode III, is thereanything you would change?
RM: Yes. The size of the 4:1 lens, which is a relatively smallzoom lens, has already decreased by half the size. We're very excitedabout the fact that we have a power cable now that's working for thestudio high-def camera. [In Australia,] we had two cables that hadactually 10 different wires in them. No different than if you weredoing huge video systems for a large major picture, but now it's alldown on one cable, and we believe by 2003 it'll be microwave-able, sothere will be no cables whatsoever to the recorder.
Remember, we record inside the camera, which is our "negative," andwe're creating a clone, but to be able to have that wireless, I think,will be fantastic.

PP: Were you doing a lot of shots with Steadicam?
RM: We did everything. We shot Steadicam. We had theSuperTechnocrane literally every single day.

PP: So that second video camera wire can really be a pain inthe neck.
RM: It's no different than having video systems in threedifferent places. But it's a pain in the ass for anybody. If theaverage film gets about nine setups a day, and that means you block outa scene in the morning, and by the time you rehearse it and shoot it,it could be an hour and a half to two hours. But when you're doing itevery 20 minutes, that's a totally different reality. And people forgetthat.
You're not going to have a huge cost savings on film if you've onlyshot 350,000 feet over a 12-week period. But if you're shootingPearl Harbor, or a film by George, where people are reallyobsessed with getting film, they need film to be able to manipulate,they're mostly "writer/director/editors."

PP: Pearl Harbor is a good example because with someof the shots, they only had one take, so they had multiple cameras setup for protection from different angles. You didn't really need to doanything like that because your [effects] shots are createddigitally.
RM: Well, it's a different kind of storytelling. What I mean is,Pearl Harbor and Titanic are still very reality-basedfilms. They go through the process of creating real things. Jim had tobuild a boat that was 92% the size of the original. That's realfilmmaking.
Because of the fact that everything was fabricated in our films, Georgetends to digitally create worlds that aren't real. We don't have to gothrough the pain of building the Titanic. That would be a CGR miniaturefor us. We shoot whatever's the most economically feasible, but a lotof that stuff that we're driving now is obviously CG work.
For Pearl Harbor, I've only seen the reel that Michael Bay didwith ILM, the 30 minutes of the actual attack itself, and it is withouta doubt some of the best work that ILM has ever done. It's one of thegreatest collaborations between Eric Brevig, the visual-effectssupervisor, and Michael because they were able to do real live-action,real planes, real boats, CG boats, CG planes, miniatures. If you wantto see the perfect collaboration between real and digital, it's thatfilm. I haven't seen the film, but I've seen the attack. And the attackis just … your jaw opens.

PP: I'm afraid the quality of the projection for PearlHarbor won't be up to the quality of the filmmaking, based on myexperience with Shrek.
RM: I know. And seriously, if you know what we go through, notonly just in terms of the image acquisition, but also what we do interms of the mixing, and to go into a theater where two of the speakersare not even working, there is no surround system, the print istraveling 30 or 40 feet over wires to get to the screen--it'stragic.
It's not just the fault of the exhibitors. It's a globalpicture--there's a lot of problems. We kind of jury-rigged thistechnology for 50 to 60 years because we never put enough money intothe actual system. Kodak's done some fantastic things with their filmstocks over the last 5, 10 years. But the problem is, they will alwaysbe hampered by the technology that projects those images. And it isn'tabout just going in and buying new film projectors. It doesn't haveanything to do with the film projectors--it has to do with theprocess.
There is no way to create 5,000 identical prints on film. You go to alab, and there is not a single owner of a lab that won't tell you thatpart of his operation is run like a Laundromat. When you're doingrelease prints--I don't even know what the speed is …

PP: It's like making photocopies.
RM: Exactly. But the high speed at which it goes and thechemistry involved means that the print you start off with, number one,and number 1,000, are so different and opposite ... tragic.
I think the best world that we can have is this: Get through the fear,the politicization, the labor issues that are involved in transformingthis huge wasteland of arenas that we have that are called theatersinto a reality where people can actually see and judge films with atotally different aesthetic, whether it's projected on film ordigital.

PP: Where do you think the biggest growth is, short-term, forelectronic cinema? There are a lot of things going on with people doingalternative cinema, independent filmmaking. Do you think that's whereit'll take off the fastest?
RM: I think everything that's really been a dramatic change inthe history of the industry has come from a small place, a place whereyou would least expect it. I think one of the great things in terms ofthe distribution and exhibition issues are players like Boeing, or GM,or any of these other leasing companies that may or may not get in.They have the money and the vision to be able to make a decision.Getting six or seven studios to come up with a standard and make adecision … well, I've never seen it in my lifetime.
I wish them all the best, but there's so many different competing andfearful aspects to where they're going. It has always been pioneers whohave changed the dynamics of how we acquire and exhibit images in adifferent way. You look at those old charts that you usually see inlabs that outline all the formats and all the different soundtracks andall the different ways that people have made and shown movies, and it'salways been developed by individuals.

PP: Is there anything you'd like to see in future digitalcinema cameras?
RM: The resolutions that we're at right now are the beginning,the most primitive steps in the evolutionary nature of digitalcinematography. Can you imagine, when we have a 10-million-pixelcamera, and two or three years beyond that, we'll have a 20-, or 30-,or 40-million-pixel camera? I've just ordered a notebook that's aone-gigahertz model, with a gigabyte of RAM. That was unthinkable fiveyears ago.
There will be a 10-million-pixel camera. I don't think it'll be readyby Episode III, but I think it will probably be done the yearafter.

PP: Have you seen the new projector that JVC's showing?
RM: Kodak was kind enough to give us a demonstration last week,very impressive, and we're very happy that they're on board now.

PP: JVC claims they're going to be able to deliver a true 2Ksystem by the end of the year.
RM: We want them out there as much as we want TI out therebecause the more people who come into this, it's only going to do onething. The competition's going to be great, and with competition, youactually begin to see results. The problem is, we've never hadcompetition within the film business.
When Doug Darrow from TI came to Skywalker Ranch about two and a halfto three years ago, we'd been talking for years, and he said, "I thinkwe're ready to show you something." George and I looked at it, I thinkin November of '98 when we finally saw the "Mark 1." And I asked, "Isthere any way, Doug, that you can have this ready?" And without even apause, he was there, and he delivered.
The minute he agreed to do it, JVC agreed to do it. We had fourtheaters distribute Episode I, and whatever anybody thinks aboutwhether they liked it or not, for both of them, those were the only twoprints in those four theaters that resembled the film that theymade.

PP: It's at the point now where you're targeting a MemorialDay 2002 release for Episode II. Do you anticipate digitalscreenings in some cities?
RM: We're hoping. This is my personal pipe dream and George'soverall dream: I hope there's 250 digital cinemas out there. There are36 TI projectors out there now. From everything we hear, they cancertainly get 250 out there. I think Kodak is desperate to get theirdigital projectors out there. They're realistic. You know, they're filmpeople.

PP: They're pragmatic.
RM: Part of their company is a digital company. But they'vewaited a long time because I don't think they actually believed digitalcinema was going to happen. And now it is, and now they've got to catchup, and we're thrilled that they're catching up because the morecompetitors that come in, the greater the different systems willbe.
No matter whose projector is used, you have to make the digital printfor the best system--you can't compromise on that. It's like the firstthing that TI has ever spoken to us about. What they're trying to do,as is Sony, is when we finish a print, you grade your answer print ontape or a server or whatever that medium is, and it goes with a chipthat goes into the projector. If the luminance of the bulb is notbright enough, the projector shuts down. Plus, it reads the lookuptables of every single frame, and interprets it so that every singleprojector is in synch to that information.

PP: Any closing comments?
RM: Can I tell you my dream? Five hundred digital cinema screensby 2002, but by 2003, a minimum of 5,000. For 5,000 [screens] spreadout all over the U.S., you would have a discerning public of which5,000 theaters could definitely reach. And it would make all thedifference to them in the way they enjoy the film, whether it's ourfilm or any film, whether it's acquired on film or digitally. Itdoesn't matter, as long as it's finely projected.