Julian Clarke, ACE was nominated for a BAFTA, an ACE Eddie and an Oscar for his editing of District 9. He was nominated for an ACE Eddie for Deadpool and won an ACE Eddie for the pilot of The Handmaid’s Tale.
His other work includes Skyscraper, Chappie, Project Almanac and Elysium. He also cut episodes of Love, Death & Robots and Altered Carbon.
Today, we discuss his work on Terminator: Dark Fate, which, like Deadpool, was cut on Adobe Premiere.
I’ve previously interviewed Julian when he edited Deadpool.
This interview is also available as a podcast, though, because of technical difficultis, only the first 30 minutes (out of 60) is in the podcast.
HULLFISH: You’ve done a lot of sci-fi. Do you feel like you’re being pigeon-holed? Or are you just happy that you’re getting the work?
CLARKE: Nobody wants to be a one-trick pony. But it becomes comfortable for you and it becomes: “Well he’s the guy who does this.” So it becomes easy to carry on the pattern. I also carry on the pattern because I like that stuff. I’m attracted to highbrow genre stuff — whether it be horror, fantasy, sci-fi, crime movie. To me, I like that stuff that’s smarter adult-oriented R-rated but immersed in genre. A lot of that stuff tends to happen in science fiction, but I’m kind of interested in that across the board. I’m also interested in comedy and drama as well. Deadpool was a fun project because it was so multifaceted. I got to dip my toes into comedy more on that one, and I think I’ll do a bit more of that in the future too for sure. After I worked on Deadpool, I worked on the pilot episode of Handmaid’s Tale, which technically called science fiction but it was almost like working on a period piece. It was dealing with the real minutia of what the actors are doing with their eyes and what’s said and what’s unsaid. So it was a very different type of editing than robots and explosions.
When I worked on Handmaid’s Take it was so nice to work on something that’s not effects-driven and is all about like real subtle stuff happening. The club of people making well-budgeted dramatic movies is very small. It’s a very small number of movies happening in Hollywood and it’s a very small club of people working on those movies. It’s a very exclusive little niche to work on the well-funded interesting dramatic movies out there. I’d like to join that club at some point.
HULLFISH: It also has to do with the directors you work with, right? Those guys are working on those types of projects.
CLARKE: Sure yeah.
HULLFISH: Something I noticed going through your IMDb page was that you’ve done a bunch of shorts. What’s the value? What’s the purpose? Obviously District 9 started as a short and became a feature. Why is Neil and why are you working on those shorts?
CLARKE: I was working on them because I like Neil and he does cool stuff and why wouldn’t I? For him, the money he’d make or the pragmatics of things really don’t play a major factor.
He got a bunch of money from Valve — the video game company — they were interested in exploring making some content and for him, it allows him to do a bunch of ideas and not have to spend a year doing them. So it just became kind of like an “idea lab.” He could do a bit of a Vietnam sci-fi thing. He could do another one in South Africa.
When you’re the director it means committing multiple years if you want to do a feature. There’s the development and the writing and then there’s the shooting and then the editing and then you’ve got to go out and tour around with it.
HULLFISH: Does editing those shorts stretch you in different ways or allow you to use different muscles?
CLARKE: They’re a lot more impressionistic. They’re less narrative. You’re don’t really need to have this beginning-middle-end thing or pay off a big character arc. You can have something that is more of a fever dream when it’s short. A fever dream that’s two hours long? People make them, but that has very something that like you know as a very niche audience. But people will watch a cool 10-minute fever dream that’s on YouTube or something like that. You’re less beholden to this sort one plus two equals three kinds of storytelling.
HULLFISH: Can you pull that stuff into your feature work? The things that you learn on those shorts?
CLARKE: Maybe in a particular type of sequence where you’re kind of doing something less linear.
HULLFISH: You’re talking about being less linear, and with this movie — Terminator: Dark Fate — if you’ve got a storyline that is based on time travel, does that allow you to change structure or be less linear in storytelling.
CLARKE: You would think that maybe the Terminator movies would be like that but it’s not Back to the Future II. Genesis was trying to become more of a time-bending movie like Looper or something like that.
But time travel is just a means to set the table, but there’s not like a lot of bouncing around. Aside from the fact that we have a couple that kind of flash-forwards that we could shuffle around in the movie since it’s a chase, it really is actually: “this leads to this leads to this leads to this” right? And so in a way it was actually kind of a movie that — even though it is time travel — it was quite linear.
The only structural stuff you could do with it you would be “where do we check in with the villain?” “Where do we put the flash-forwards?” Those were the elements that could be moved around. And then the rest of the stuff is just what you could move around within the sequences — like getting out earlier.
Deadpool was a much more structurally-malleable movie than this one was.
HULLFISH: So you’ve kind of jumped back and forth between directors you’ve worked with several times. Do you have to get into “Neil mode” or “Tim mode” when you switch?
CLARKE: I’ve worked enough with both of them that it’s in the muscle memory, so I can just slide right into it. They’re also not a radically different way of working.
Tim likes to sit in the room a lot, though he didn’t get to sit in the room as much as he did in Deadpool on this one because of the sheer quantity and scope of the VFX work. He would just be in six hours of VFX reviews a day. So he didn’t get to hang out with me as much as he likes to hang out with the editor. Some directors are a bit too fidgety to sit in the edit room for too long. It’s all a bit too glacial for them — the process.
But Tim likes the glacial process and going through the minutia with you when he has the time, but he didn’t have quite as much time to do that on this one. I’m sure he would’ve happily edited it for several more months and gone through the dailies with me more. I’m gonna get it
HULLFISH: In The Conversations, Walter Murch mentions that each director has specific foibles or things that they hate, like, for example, cutting in the middle of a word, or wanting to wait for action to stop before cutting. What are some rules that Tim and Neil have that you need to remember?
CLARKE: Tim Miller hates double-cut action, which is tricky because when stuff happens really fast, double-cutting is something that you often go to just so you can kind of slow the action down long enough that the shots are long enough that you can kind of read what’s happening. But Tim really likes it to not be double-cut. Sometimes he lets you get away with a little double-cut, but VERY double-cut — like you have the thing explode three times in a row — he’s not a fan of that.
HULLFISH: Murch mentions a director that hates pre-lapping dialogue over the establishing shot of the building, like “What? Is the building talking?”.
CLARKE: Yeah. I’m not a big fan of pre-laps either but they sure help you speed up the pace or if a transition feels kind of sloppy, they can smooth it, so sometimes I do it, but I don’t love it.
HULLFISH: You’re back on Premiere again. Are you on Premiere for good?
CLARKE: I’m not a disciple of one platform. On Deadpool, Tim said, “We’re editing on Premiere,” and I was like, “Okay. Cool!” There were a lot of challenges to that one, so once we did it we talked to Adobe and told them, “These are the things that would be great to work on.” Tim’s definitely sticking by Adobe for sure.
So when we came on to Terminator Adobe had basically made a whole new version for us — which I think is probably going to be publicly released pretty soon. They basically did a bunch of stuff that lets you do kind of Avid Unity-style networking and shared project stuff. Basically, it’s like bin-locking. It’s not a very sexy feature because it’s been around for a long time with Avid, but it is unique in that the other NLEs haven’t bothered to make that happen, and it is integral to a movie of this size that you have that functionality, because we had four assistant editors, a VFX editor with two assistants, so we have eight or nine people like in the project file updating VFX, and pulling lists.
On Deadpool, we were doing it all at the finder level. We got away with it on Deadpool, because we were smaller, but this one was had 2300 VFX shots. We really needed that functionality, and it really worked. So even though this movie was a lot bigger than Deadpool it actually was a much smoother process. And Adobe’s going to work on it and evolve it further, so on the next Miller project, it’ll be even more souped-up.
HULLFISH: The last time we talked about using Premiere, I remember there were issues and we talked about what some of the challenges were but that was quite a number of years ago.
CLARKE: On Deadpool, that was 2015.
HULLFISH: So in four years it’s come a long way. Would you say bin-locking or project sharing is probably the biggest thing?
CLARKE: I told them to essentially focus on the non-sexy stuff because really the non-sexy stuff is often the stuff that causes your assistant editors to work, till 2:00 in the morning, right?
They worked on getting the project files to open way faster. On Deadpool, we had really big project files and they would take like four minutes to open and now — because of the combination of working on how fast they open with the way that they are now splitting all these things in separate projects — essentially a bin is a project — so now they open up in 10, 15 seconds.
That combined with the networking and then having it integrate more successfully into ProTools so that turning over to sound was a much smoother process. There’s still a little bit more work to do there in terms of change lists.
Across the board, all that not particularly exciting having-the-whole-machine-churning-efficiently-and-non-frustrating stuff advanced light-years,
HULLFISH: So you really are using essentially a different project for every scene?
CLARKE: Yeah. If you’re an Avid user, think of a project as a bin. Then there’s a new type of project which holds all those projects: a master project. So you create a project file for scene one for scene two and a project file for reel one and two, et cetera. So it’s a project within a project basically.
HULLFISH: Got it. Is there anything that you really find that you love in Premiere that when you go back to Avid you miss?
CLARKE: There’s a lot of stuff in Avid where I use it and it just like feels ancient like audio suite makes me want to kill myself like you know Dever but the energy was like he was like like Why the hell is that thing still around right. I mean the way just like that you say, Okay, here a filters tab and here’s a stack of audio filters. That makes way more sense. It’s just simple. Intuitive. AudioSuite is horrendous. And the same thing about stepping in for using filters. All that stuff is so clunky and slow and feels like it was designed like 20 years ago — which it was.
Avid’s sort of the beat-up car that keeps running forever and will get you there, but there’s some pretty creaky stuff that’s been in it for a long time.
HULLFISH: Let’s talk a little bit about your approach. Has it changed any since the last time we talked? When you’re looking at a blank timeline what’s your approach to viewing dailies and then starting on a scene?
CLARKE: It kind of depends on how much has been shot and how much time I have. If you have five hours of dailies you could watch all of them and then start cutting and you’d basically be nowhere because you’d sunk a huge portion of your day into watching everything. If they’ve shot a little bit then you watch it all or if they shot a ton, then you kind of have to prioritize — what’s the stuff I’m going to watch all the way through? Watch the stuff I’m going to kind of scrub through? So it’s kind of just time management in terms of what you’re up against.
It’s a nice luxury to get to watch everything, but sometimes watching every single frame is not possible based on the quantity they’re shooting. If you have multiple editors then you can split it up. But I was on my lonesome on this one.
HULLFISH: If you listen to many of the Art of the Cut interviews, people will say, “I watch everything EVENTUALLY.” When you’re cutting dailies, it’s a different story.
CLARKE: When you come back and you’re revisiting the scenes, then you definitely finish seeing what you didn’t see in the first go-round.
HULLFISH: So, you were saying that you have to decide what to watch completely during dailies. What do you watch completely? What do you scrub through? Glancing at wide shot masters and diving into the closeups?
CLARKE: Yeah. Wide shot masters are often — depending on how the director works — that’s often where he’s kind of warming up the actors. So if they did 15 of those I might not watch all of those. Often the actors are saving their good stuff for the close-ups as well as the fact that sometimes the director’s trying to find it in those first masters. Sometimes you can find some stuff that is quite different in those early masters that is interesting but more often than not they’re more disposable. I’m definitely going to watch the circle takes, and a bunch of the close-ups probably. Less of the masters and if there’s boring stuff like inserts of phones and stuff like that, that’s something you can definitely scrub through.
HULLFISH: Once you know you have the insert of the phone, that’s all you really need to know.
CLARKE: Yeah. And the close-ups are interesting because often the last takes are the circled takes and those are the ones that are closest to what the director’s vision is. But then sometimes those are ones that are the most polished in a way that they can be slightly artificial too. So sometimes you find the earlier takes are more off-the-cuff and they’re more naturalistic. They kind of hit the beat harder than the later takes. So that’s a toss-up of whether you want a performance that sells it really hard but has a little bit of artifice to it or something that’s kind of more naturalistic than the earlier takes.
HULLFISH: Do you learn the working habits of your actors? Like, Ryan Reynolds is always great on take one and two but by take 12 and 13 he’s bored but his co-star is always rough on takes one and two and is on point in the final takes.
CLARKE: There are certain actors who are very consistent and then they modulate a little bit, or they improvise, so you have to watch everything, then the circled take really means that the director prefers this improvisation the most. And then there are other actors that are totally erratic and you have to kind of piecemeal the performance together.
HULLFISH: When you start piecing that stuff together, do you use selects reels or are you just going through bins or do you put locators in? How do you remember all those great moments?
CLARKE: I find with action stuff, they do it so many times and it’s often very quickly cut and so when you’re trying to figure out what’s the best punch, then I’ll get my assistants to break it down to each beat. So here is that punch from every single camera angle and take all in a run and then you can decide, “This is the best. This is the best. This is the best.” And then even make a mini-string-out between those and then pick and do that for all the action stuff.
That gives you a good starting point for finding your favorite pieces because the action stuff can kind of be overwhelming with how much they shoot and how many angles — and sometimes the differences between the takes are very subtle: like this punch is 10 percent more credible than this one. The ability to decipher that from memory is something I don’t have. So the string-outs are kind of essential with that stuff.
Then with the dialogue stuff, It really becomes more of a question of how much they shoot. I can kind of hold the dialogue performance stuff in my head if it’s not a horrendous amount of takes, but if it’s a huge amount of takes then you need to start putting in locators or stringing that out as well.
I kind of think it would be great to have string outs of the dialogue stuff as well, but I’ve kind of felt like that’s inflicting a lot on the assistant editors to do that on top of the action string-outs, so I only ask for it when I’m desperate. It is actually very helpful. In some scenes, like where the director is resetting a lot or doing a lot of alts or jumping around in the scene, then it really becomes very hard to know where stuff was in your bins and then stringing it out is kind of helpful because when they’re not making a take be a take, and they’re jumping around a scene in a given take, then locators only help you so much. It becomes pretty confusing and you can be pretty inefficient in finding stuff, so string outs can definitely help that kind of situation.
HULLFISH: How are you organizing things in bins? Do you use the same workflow or the same methodology in Premiere as when you’re working in Avid?
CLARKE: Actually, yeah. That was another feature that Adobe made for us. I don’t think it’s been generally released yet, but they basically made a custom bin setting, so you can move stuff around like in Avid instead of having this retiling thing that has been the Adobe standard. So, I like to do that layout where you’ve got an A cam and B cam and C cam below and then the group clip of all of them on top. Then you have your masters and then your close-ups and then your inserts at the bottom.
HULLFISH: I haven’t seen that feature yet.
CLARKE: That might be part of the next release.
HULLFISH: The Premiere folks I spoke to before this interview said they’d done a bunch of new features just for this movie.
CLARKE: That was one of the ones they did for me. Well, me and probably a bunch of other editors.
HULLFISH: Oh I’m sure of that.
CLARKE: When your NLE has your organizational system be very intuitive and well dialed-in, whether it takes five seconds or 30 seconds makes a really big difference in terms of how the vibe of the room is.
HULLFISH: I love that idea about how important the vibe of the room is, because — as good of an editor as you can be — if all of a sudden something goes wrong in that room that’s not your fault necessarily — it makes a difference in the way your work is perceived.
CLARKE: Absolutely. The negative energy in can be brought in to how people look at your work and in terms of how you’re pitching your ideas and stuff like that. The psychology of how people perceive things is weird. For a long time I had a don’t pitch attitude when you’re doing changes, because sometimes people kind of talk themselves out of the new idea before you even show it. They think, “That’s not going to work because of this, this and this.” Now you’re going rogue by doing it. So it’s better to show them first than to talk about it and pitch the idea and THEN show it.
HULLFISH: I really am fascinated by that concept because I’ve heard different people say different things about how they pitch a new idea or when to pitch a new idea. So let’s talk about that a little bit. Especially in an assembly, if you’re cutting something that a director is going to see for the first time, do you make sure that you deliver it as the script is, or the way that you think is best? Or do you deliver it in a way that you think the director would want it?
CLARKE: I think you need to have a read on the director’s personality, because if the director is going to be horrified by you deviating, then you take that on board. And if they’re not going to be horrified, then you kind of feel a little more free to experiment.
I definitely feel obliged to try to execute the plan. Sometimes it feels like the plan is not going to work. And then you can show them a version of the plan that works. I try to follow their approach as closely as I can.
For instance, in the scene where Grace dies, in the script there was a very different order to when the Terminator comes out, and when it sees them. In the original, scripted version the Terminator came out way earlier and then it was kind of meandering around, not finding them. It felt like the tension was inert. I felt like it needed to show up and immediately kind of go for them. So I cut it that way right off the bat because otherwise it felt really wrong. We edited that in a million different ways, so it kept evolving, and Tim was fine with that.
SPOILER ALERT OVER
HULLFISH: Usually the problem with NOT doing the plan, and just going straight into your own idea is that you’re not giving the director a chance to understand that his way won’t work.
CLARKE: I definitely like to show the director what their plan is. If you’re going to leave Plan A you’ve got to kind of carry them along with you. Or if you have the time, you cut multiple versions. You cut the Plan A version, then you cut the Plan B version. Then it gets complicated because eventually those start having different structural destinies as well and then you don’t just have two different scene versions, but two different structural cuts of the movie.
You want to spend as little time as possible on multiple structural versions of the movie because it just becomes really labor-intensive to update the VFX and stuff like that. When you finally get into the director’s cut, are you cutting in reels? Or are you cutting single scenes? Do you attempt to have the entire movie, like a screener? No. On any NLE I’ve ever used, the performance is terrible if you have the whole movie.
Despite the fact that we’re no longer projecting 35mm, reels become a functional way to break up a movie. Once I have a third of the movie shot, I start to build the thing and put in cards where the missing scenes are, so I start to build the structure as it’s being shot. If you’re looking at it purely on a scene level, you’ll miss things like pace or structure or transitions. This big-picture stuff that you can catch if you’re building the structure as you go, whereas if you’re just dealing with it on a scene level, you’ll miss it.
HULLFISH: That happened to me on a film that I cut recently where I cut two scenes separately but while we were still on location I put them together and realized that the two scenes needed a transition shot to bridge them. If I hadn’t put them together, I wouldn’t have know that something was missing.
CLARKE: It can be a great scene on its own, but that doesn’t mean anything. How does it behave in the context of the movie, right? There’s a lot of stuff that kind of changes once you kind of have it all in a row. To me, that’s the whole benefit of having an editor assembling as you’re shooting. Hopefully, there are not too many of those things, but there can be little things you can grab.
HULLFISH: Any specifics on how you’re collaborating with the director as you’re going through the director’s cut?
CLARKE: We screen it fairly regularly and he loves to show people so we’re always pulling people into our screening room. We’re always generating feedback. So there’s always something to work on. Despite coming from a VFX-animation background, he really is like an actor’s director. He is very focused on performance. He has an incredible BS detector. When something seems a little bit artificial or forced or whatever, he can sniff it out.
HULLFISH: You mentioned the screening room. Did you actually have a nice little multiseat projected screening room? And how were you monitoring audio coming out of Premiere?
CLARKE: We monitored in 3.1. I think I did that for the first time on Skyscraper. I haven’t done 5.1. I feel like it might be a little too much effort. I love having really great sound. 3.1 lets me push the music off the dialog and have a kind of a boom track that just makes it feel much more like an action track, cinematic sound. And so you can definitely create a much better, immersive sound experience. I’m all about trying to simulate the finished product as much as possible. 3.1 is enormously helpful in terms of doing that.
In terms of our screening room, it was basically a little space that we shared with VFX. They brought in a nice projector that they used for VFX reviews. It wasn’t massive. It could probably fit maybe 15 or 20 people.
HULLFISH: So for those who have not heard the term 3.1, that’s really like LCR with a subwoofer.
CLARKE: Yeah, exactly. Yeah.
HULLFISH: Did being in that screening room and watching screenings affect you in any way differently than being in your cutting room?
CLARKE: Well, for sure. Wide shots play so much differently on the big screen. I think you notice different things between the wide shots and the close-ups. When I watch the movie in IMAX there are certain edits that I feel could be longer now because there’s something about having it fill up your whole field of vision and it becomes more overwhelming. It takes slightly longer to fully take in all the information. So then, what screen size are you perfectly calibrating your edit for?
For sure it’s useful seeing it on a different size — closer to movie size. I wish I could screen it on a REALLY big screen.
HULLFISH: Another thing I thought of when I watched a big VFX movie that probably had a ton of shots is that when you’re cutting those shots in with previs wide shots, they probably aren’t very interesting or compelling, but then when you get the finished VFX shot in, you probably wish you’d cut it in longer because now it looks so amazing and you want more time to take it all in.
CLARKE: Yup. I learned that on Deadpool. You can read pre-vis really fast because it doesn’t have motion blur or particles. But once you add those things, then the shot becomes infinitely more complex to read, and it almost needs to be double the length. Shots that I had cut into action scenes at 10 frames needed to be lengthened to 36 frames because the motion blur would just completely transform your ability to comprehend what you’re looking at. So now whenever I think of cutting in an incredibly short previs shot, I make sure to pad it.
HULLFISH: Also the previs doesn’t have textures and there’s no facial expression or emotion.
CLARKE: Also, you can do amazing, impossible camera moves in previs and it looks really cool and dynamic. But if you do that and you see the final shot, it’s like, “Oh! CG camera” when you view it in high res. It just seems fake. It’s too cool and perfect and so you actually need to make the camera work sloppier. So, when you do the effects, you need the camera to overshoot and pans back. You need to intentionally have imperfections to ground it and make it seem like a real camera.
HULLFISH: When you were talking about editing dailies and delivering them to your director on set. Were you doing that through something like PIX? And were you getting feedback during the shoot? I’ve talked to several editors who deliver scenes to the director to view during production and they’re just too busy to respond.
CLARKE: I think Tim loved watching those scenes during production. He was so excited to see a cut scene of something that he’d shot two days ago. And some directors — when they get those scenes — they hoard them… they keep them to themselves, but Tim would show them to everybody, like, “Hey, craft services! Look at this!” As for actual feedback I only get that erratically. This was such an intense shoot, so even on his days he was having meetings and was engaged in other stuff.
HULLFISH: A director I worked with would bring in a few members of the crew — mostly camera crew — and show them stuff at the end of the day. And when he figured out that I could Airdrop the scenes from the Avid onto his phone, that was like a miracle. There’s probably some I.T. security company that’s flipping out.
CLARKE: Yeah. Our editing computers weren’t even connected to the internet. It’s interesting for you to say that the director came in for you. So often, the director might be in another country and they’re just sharing cut scenes over PIX. But they like the director to actually come into the cutting room so that they can show them — they can react AND the director can’t re-watch the scene a thousand times. It’s much more like watching a movie in a theater where you see it once and give your honest, initial reaction instead of picking it apart by watching it over and over again. You don’t sweat the small stuff.
HULLFISH: Tell me a little about that shooting experience. They were shooting where? You were editing in L.A. and using PIX to get dailies?
CLARKE: Yeah. We would send dailies and cuts that way. In the old days, you’d do a DVD burn, but that was a nightmare because you had to put a unique burn on each DVD. (Adding a watermark of the name of each person over the image.) That’s a full-time job for one person. So PIX adds the watermark to the stream and you only have to upload the one version.
HULLFISH: And was your team getting dailies via PIX?
CLARKE: No. We were getting Aspera downloads. I think the lab was delivering the dailies to the studio with PIX.
HULLFISH: On my last bigger film, we were using PIX to actually deliver cut scenes to the studio instead of all of the dailies. It was a faster way to show that the work was getting done and the progress and quality.
CLARKE: Hmmm. For me the cut scenes were only going to Tim and if he wanted to show people, those cut scenes were only going to the people he wanted to show. I wouldn’t be sending cut scenes to the studio.
HULLFISH: The director definitely has to make that decision. The director doesn’t want the studio to see something that they don’t approve as a cut scene.
CLARKE: Yeah. Exactly, especially with a cut scene, the editor’s cut of a scene might not be exactly what the director wants it to be, and then you’re entering into the politics of what the movie should be very early on. That shouldn’t happen until after the directors cut.
HULLFISH: I wouldn’t send the cut scenes on PIX until the director approved the scenes and the studio definitely knew that it was just the editor’s cut. But for upload and watching purposes, it cuts the amount of footage that gets sent to the studio from maybe six hours a day down to 4 minutes.
CLARKE: If the director is on really good terms with the studio and feels really confident in their power I’m sure that would work great. Information is power. So the question is: does sharing the information in this way empower the director or not. It probably really depends on the specific film and relationship.
HULLFISH: True. The relationship would count. And this director had absolute final cut.
CLARKE: So in that case, there’s nothing to lose. And with a cut scene, the studio could air their concerns while there’s still time to do something if the director decided to listen. That seems like an ideal situation.
HULLFISH: Do you remember how long the assembly was? And then where you got down to for the final movie?
CLARKE: It becomes a bit approximate, because there were some scenes that were really long and on those, they were cut down as part of the assembly. So those scenes do exist as longer versions, but they never made it to the assembly. If we counted those scenes, then the mega-cut would have been about three hours, but because I started slashing and burning during the assembly phase it came in around 2:43 as a first cut. Then the director’s cut got down to 2:20 and by the end of the process it was about 2:05.
HULLFISH: Were there any big challenges in cutting this film?
CLARKE: The biggest challenge in this movie was in handling the protagonist. In the first movie you had Linda Hamilton then in T2 you had Linda Hamilton and Arnold. And in this one we have Grace and Dani and Linda and Carl, so we have four protagonists plus this alternate Skynet storyline. So in a way we’re doing an alternate villain as well, so it becomes a whole lot of information and character arcs, so finding the right amount to feel like you were servicing each character arc and have it feel real and satisfying without having the middle of the movie bog down.
In a way, we were kind of emulating T2 in a way which had this dramatic middle section followed by a relentless third act, so we kind of mirrored that structure, but with so many characters, it was kind of easy to bog down in the middle. We didn’t want to feel like we’d left the action movie behind for too long. That became the major challenge of the movie.
HULLFISH: Did that take a lot of full-length viewings of the movie to get the sense of that? Or could you sense it on a more micro-scale than that?
CLARKE: Well, I think both. The hardest character was the Dani character because she changes the most. She goes from being this naive innocent to being on the road to being this future leader character. That’s sort of similar to the Linda Hamilton character’s journey in the first film, but in a way, this needs to go even further because in the first movie Linda Hamilton was just the mentor of the future leader. And on top of that expansive character arc we have to give Linda and Arnold their due since they’re franchise characters. So giving Dani the space to grow into that character was actually quite delicate work.
HULLFISH: As many other editors I’ve interviewed have said about other action films: the action doesn’t really work if you don’t care about the characters.
CLARKE: Yes, and you also have to have that character stuff in the action scenes too. To me, the first half of the movie is a little more effective than the second half in that regard. In the freeway chase, I feel very anchored in the characters. I feel very anchored in the characters in the detention center. Those are scenes that are exciting and dynamic and you’re also very connected to what the characters are feeling and the jeopardy.
That becomes a lot more challenging in the scenes with the T5 and it becomes so VFX-driven and spectacular. That’s where it’s harder to keep your hooks firmly tethered into the characters.
HULLFISH: You were talking about the process of getting yourself from an almost three-hour long first cut to an almost two-hour long final cut. I’m assuming that was a combination of big macro cuts and also finer trimming of each scene?
CLARKE: Yeah. There were a couple of big lifts and then a lot of snipping out dialog and stuff like that. It was definitely the right way to do it, but a lot of scenes were a bit over-written in the middle because they wanted to explain how things worked but it was just too much, so we had to kind of boil and reduce the scene into the essential information along with a thread of character. There was a lot of that that happened in the middle of the movie. Then there was a big chunk that came out quite late in the editing process in the border-crossing sequence. So we end up doing a pretty drastic reduction on the border crossing sequence. And so it became a little more transitional. It used to be a set-piece, but it didn’t kind of measure up to the other set pieces.
HULLFISH: It all depends on the movie. Sometimes it’s chipping away at smaller pieces and sometimes it’s more structural. Sometimes the scenes feel like they’re as tight as they can be, so you have to go for big chunks.
CLARKE: Sometimes you realize that there are whole scenes that are kind of redundant. Maybe it’s a beat that reinforces something that’s already been said. No editor likes it where you feel like the scenes are great but the overall pace of the movie is bad. So then you’re yanking out content of a well-paced scene just to kind of get the overall number down. In those cases, it’s nice to find a whole lift that can go so you’re not overly tightening the dialogue or pacing of good scenes.
HULLFISH: Or sometimes, it’s even an awesome subplot or B story, and as cool as that subplot is, it’s got to go.
CLARKE: In this movie, there was a piece with Linda Hamilton after they stole her car and she’s left with the rev 9 reforming. There was a scene of her stealing another car. She orders a guy out of his car and takes his car and drives off. Then there’s a section where she surreptitiously follows them. But eventually, you realize that you don’t really need that. You know that Sarah is resourceful warrior. She just shows up with another car? You get it. She’s not going to be stopped by having her car stolen.
HULLFISH: And did some of those friends-and-family screenings or more formal screenings prove that out? You could see that a general audience could make the leap?
CLARKE: The thing about friends-and-family screenings: you can never trust them. You can never trust the general enthusiasm level, but what you can trust is confusion. You can definitely tell if they’re not tracking. You can get some sense of pacing, but they’re generally pretty charitable with pace as well. That’s something you definitely want to gauge before you deliver a director’s cut because you don’t want to do a bunch of big lifts and feel like you’ve created something that’s incoherent. So for sure, we’re testing out those experiments in screenings.
HULLFISH: Got it. Julian, thank you so much for your time today. Really appreciate talking to you.
The first 50 interviews in the series provided the material for the book, “Art of the Cut: Conversations with Film and TV Editors.” This is a unique book that breaks down interviews with many of the world’s best editors and organizes it into a virtual roundtable discussion centering on the topics editors care about. It is a powerful tool for experienced and aspiring editors alike. Cinemontage and CinemaEditor magazine both gave it rave reviews. No other book provides the breadth of opinion and experience. Combined, the editors featured in the book have edited for over 1,000 years on many of the most iconic, critically acclaimed and biggest box office hits in the history of cinema.