THE MAN WHO FEELS NO PAIN

A New Process For Action

It was early 2016 when director Vasan Bala reached out to me. Having directed an amazing short film and a critically acclaimed feature film called Peddlers, he had a new idea for a Jackie Chan-style, martial arts action comedy. After seeing my short film Rope-A-Dope, he contacted me with this:

“I want to do Rope-A-Dope action in a Bollywood film.”

I read the script for his comedy Man Who Feels No Pain. It had just the right mix of comedy and drama, with enough interesting characters to make an action director jump at the opportunity. I asked him,

“Are you sure you want action like Rope-A-Dope?
Action like that is a very involved process.
No wires. No trick photography. No stunt doubles.”

He said,

“Cool, cool. We can, we can.”

When directors ask for action like Rope-A-Dope, they usually have no idea what they’re asking for. It’s like a neighbor asking for your French toast recipe. You don’t just give them a list of ingredients; you tell them what kind of cast iron pan to fry it in, the oil to use, the kind of seed bread, the kind of fig spread and what sausage to pair it with. Then they’ll throw it all out, cook their 24-hour-diner-grade French toast and complain about your recipe. They don’t realize that the recipe for great things is in the process.

In the action world, action works the same way as French toast: When we, the action team, create action like Rope-A-Dope’s, we design the action for the actors. We have total control over the story of the fight scenes, production design, prop creation, camerawork, editing, and sound design.

This is vertical Integration.

Vertical Integration is how Charlie Chaplin created the best comedies, how Jackie Chan made the best martial art films, and now it’s how we create our action. Vertical Integration in action films (and comedies) is fragile by nature. All departments have to be on the exact same page, stacked up like a tower of blocks. If one block is out of line, the structure collapses. For example, your bar brawl may be cast with amazing stuntmen ready to hit the concrete, but if the cameraman doesn’t frame it properly, the action collapses. Your kickboxing fight might be shot to perfection, but if the editor cuts at random because he doesn’t understand action, the fight is ruined. If your kung fu fight scene is perfectly performed, shot, and edited, but the sound designer replaces the block sounds with light pats and removes all the miss sounds, the fight is depleted of all its power.

Shoddy fight scenes manage to find their way into movies all the time, and the audience will often give them a pass. Who’s going to let a badly-shot or poorly-edited fight scene ruin an entire film? The audience won’t be able to discern exactly what’s wrong with the generic action, but they won’t remember anything good about it either. They watch the film or show, and the producers, distributors, and sales agents see a sale, so they keep the process the same for the next project. If it ain’t broke…

From an organizational standpoint, it makes sense why films and television do not vertically integrate the action: the filmmaking pipeline is compartmentalized for the sake of economy. It’s like a car factory. Each department of the company does its job and doesn’t dabble in other departments. Workers are plugged in and removed at various times during the operation, things run smoothly, and the model makes money. That’s what shareholders want.

Here’s an example of how the standard filmmaking pipeline plays out with action design:

• The fight coordinator is concerned with doing his job safely. The camera and editing teams are crucial to the proper depiction of the fight, and his previz might have camera and edit notes, but he’ll ultimately have to trust production with those things because he might have little to no communication with those departments.
• The camera department is hired to get enough shots to cover the fight. Usually the action team has limited control over camera, unless the team is the 2^nd unit itself and can bring their own camera operator.
• The editor might not understand the action that was shot, and he’ll be given so much coverage he’ll feel obligated to use it all, even though the action might benefit from less cutting. He might make use of multiple angles for effect. Rapid-fire editing, a common practice since the emergence of the forgiving non-linear editing (NLE) systems of the 1990s, might be employed to achieve an “effect” of fighting, rather than “showing” fighting.
• A producer might demand a change in the edit to showcase more close-ups of the star during a fight, confounding an otherwise decent fight scene in the interest of keeping viewers engaged so they don’t change channels or switch streams.

Given these constraints, action teams accept their role as a compartmentalized department, ultimately un-integrated with the other departments in question. An action team, in trying to ensure their vision is carried out end-to-end, might try to alter the process. But since this is a process that “works” according to the sales figures, they can expect an uphill battle. While they are always busy developing and previzzing new action concepts, they have little option than to let the compartmentalized filmmaking process do whatever it does.

When it came to Man Who Feels No Pain, we took a risk and pitched the vertical integration package we used in Rope-A-Dope, Blindsided, and hundreds of award-winning short and feature films we had done previously. The package integrated the action with script, casting, props, production design, direction, camera, editing, and sound design. We explained that, without vertical integration, we would not be able to deliver the Rope-A-Dope-style action Vasan wanted. Expecting to fight the same uphill battle, we were shocked when Vasan and RSVP responded:

“Cool, cool. We can, we can.”

It was a moment of victory and celebration. We soon realized the scope of our responsibility with the action. We now had to deliver on all the fronts we had promised. We were in charge of the action, end-to-end, of Man Who Feels No Pain. We had done it with our own films, but doing it in Bollywood… that was going to be a different animal.

PRE-PRODUCTION / PREVIZ

I started the pre-production process and hired Dennis Ruel as my fight coordinator. Dennis is a longtime colleague who also knows action end-to-end, and he was the costar of Rope-A-Dope. I asked that we be flown to India for 3 weeks to previz the action there with the actors and team so we could tailor the action to fit their skills, but Vasan was confident that anything we made in America could be executed in India. So we had to create the best action we could and pray that the actors could perform it. Having worked with many actors who fancied themselves action heroes but couldn’t throw a mid-level kick, I questioned every move we put down. Nonetheless we continued forward.

With our stunt team The Stunt People we negotiated with RSVP for a month of prep time in Oakland to prepare 4 fight scenes. We’d use previz to develop and showcase these scenes.

Previz is the process where we “block out” or “pre-visualize” scenes before shooting. “Blocking tapes” had been around since home video, with the Karate Kid team famously blocking the entire film out on video before shooting. With the advent of CG, pre-visualization (“pre-viz”) emerged as a process to block out visual effects for post, which utilized a very “integrated” technique to ensure departments were all aligned so that the CG would look right.

As stunt coordinators became familiar with the process they used it to block out fight scenes so they have a bible during shooting. In the early 2000s it was common to shoot a basic blocking rehearsal on video just for reference using a tripod, with no editing. The action team behind The Matrix pre-vizzed all the action scenes in a vertically integrated manner, resulting in a final product that looked almost exactly like the previz. Ten years later, 87Eleven Action Design, which had been birthed from Chad Stahelski’s work on The Matrix films, had reinvented the previz process in films like Wolverine and Serenity. These previz videos were totally vertically integrated, shot like short Hong Kong action films complete with set design, props, costuming, and post-production processing like sound design, visual effects, and color correction. It totally changed the pre-viz game. Today the expectations for pre-viz are even higher, as teams are expected to churn out high-quality action short films in under a week to justify their contracts, with film producers being accustomed to this level of high detail.

Having previzzed on Black Panther, Altered Carbon, and A Good Day to Die Hard, I was ready to do whatever it took to make the best previz possible.

We received the notes for the fights from Vasan:

“There’s a fight between the leading lady and four or so boys around a car, followed by a short fight between our one-legged hero and a few gents in a gym. Then, we have a fight between the three leads and fifteen opponents, followed by the final battle between four leads and twenty opponents.”

My jaw dropped. Hiring twenty people to do a previz like this would drain our previz budget in under a week. I said,

“Vasan, I can’t do a Hollywood-grade previz for this.
That’s 45 minutes of action to shoot with 15-20 performers.
I can only afford four or five of us
running around in front of a tripod,
swinging foam sticks,
blocking out choreography…
This will have to be a reference previz only.
As long as we are vertically integrated during the shoot,
this previz will be fine.”

Vasan replied,

“Cool, cool.”

I budgeted the previz so we would shoot for 4 weeks and hire 2 extra people regularly every week. We’d shoot everything we could with this skeleton crew Monday through Friday. Then on Saturdays, we’d hire 5-6 people to accomplish complex shots with multiple opponents. Our stunt team, The Stunt People, were a bunch of grown men with families and jobs, so we juggled our schedules. But the guys were excited to do some action work again, and they were all just as capable as they were 15 years previously, only now loaded with even more ideas. Matt Lucas, a long-time Stunt People member, generously allowed us to use his school The Open Matt in Oakland for whatever hours we desired. The whole team had come together to execute the best previz possible.

I started with basic blocking previz with Eddie Ray Johnson using a tripod. This was more like design-viz or D-Viz, a placeholder to fill out later with more people and more concrete choreography. We started with the first big fight in an office (fight #3). Nobody had set roles: if you had the red jacket, you were the hero; if you had the crutch, you were the Karate master.

We didn’t have photos of the location yet. I’d ask Vasan,

“What’s in the location? What’s the furniture? The layout?
The scene in the script is 2 lines long and says,
‘Fight in the office.
‘Fight goes downstairs.’
We don’t know the location!”

Vasan said there was no location yet. He wanted a pistol in there somewhere. That’s it! We had no idea how or where, so we had to invent something. Those were all the notes we had. We were building this scene from the ground up, and we’d have to take a lot of liberties. So we just went for it. We put a guy over there peeking around the corner, a guy at the elevator reading a newspaper, two guys in the other corner who don’t want to fight, and a mean guy with a stick at a desk. We just threw it all down, like a punk band making songs in a garage.

The first draft of this previz was a mess. Eddie played fifteen different characters wearing different baseball caps (we ran out of hats), the whole thing was cluttered with text notes, and most of it was shot wide on a tripod. But we had set the expectations properly, and Vasan’s reply was:

“Cool, cool.”

We kept moving forward with the director’s blessings. Our daily schedule was:

1. Drive 2 hours through traffic to Oakland
2. Coffee
3. Skype call with Vasan at The Open Matt
4. Script discussion
5. Watch YouTube videos for inspiration
6. Physical warm-up
7. Mental warm-up (choreography, drills, playing around, etc.)
8. Lunch
9. Intense, productive choreography previz
10. Pickups
11. Drive 2 hours through traffic to get home
12. Edit previz
13. Deliver previz to Vasan and RSVP

CHARACTER DEVELOPMENT

When pre-vizzing the office fight, we were just beginning to figure out the characters. The script guided us. We knew Surya couldn’t feel pain, so we could give him a cocky, devil-may-care fighting style. Karate Mani had one leg and was always on a crutch, and we opted for a Fighter in the Wind style of Kyokushin fighting for him.

As for the leading lady, Supri? The others were superheroes, but Supri… was she a Karate expert? An acrobat? What was she? Based on the script she was vulnerable, determined, smart.

So we made a scarf for Supri, a feminine shawl with sharp, metallic charms at the end, which she could use for throwing, swinging like a chain whip, or wrapping around her fists. Vasan responded with,

“Cool, cool.”

This colored her choreography for the remainder of the previz. We colored Surya with more of a Bruce Lee flavor, with Dennis usually performing as him. We created a gym in the office, with music too loud to hear the fight raging on outside, where the heroes enter and are faced with an onslaught of flying weights and dumbbells. Then we took another liberty by building a weapons lockup, where an old man meticulously signed weapons in and out, requiring that a sharpshooter villain go through a heap of paperwork just to get the pistol and fire one shot.

If we didn’t have enough people to shoot a scene, we’d split-screen. We were having too much fun creating and stopped caring about applying the typical Hollywood previz sheen, and Vasan always gave his approval of,

“Cool, cool.”

This first previz topped out at 12 minutes. This was a huge, beefy fight that spanned multiple floors. We had a workflow and we kept going.

We previzzed the final fight scene, which played out like a tournament in the script, with multiple 1-on-1 fights one after the other. This made it easy to shoot with a limited team. We all took turns playing opponents for one another. Ed Kahana showed up, Roy Chen helped us get reaction and fall pickups on Tuesdays, and Ray came to help us with Escrima moves. We had a full-blown previz workshop running. After another week, we had a 22-minute final fight scene pre-vizzed.

The rest of the fights were smaller in scope. We tried using a real car for the car fight, shooting in Matt’s parking lot, but people came off the streets of Oakland. Last time people came off the streets during a shoot, a kid pointed a gun at me and stole our camera, so we took it into the gym and set up 5 chairs as the “car”.

After finishing the last scene, we finished editing the pre-viz and exported a 45-minute long action video. We threw sound effects and a temp music track onto it for good measure, delivered the pre-viz, and got ready to fly to India.

TALENT

I had gone to Mumbai previously just before we began previzzing on a recce to meet Vasan, the cast, and some of the stunt team. It was September. Prateek brought in 10 guys to audition for various roles. One stuntman I’ll call “VT” was introduced as an industry pro. I said, “Show me a front suicide.” He landed on his butt. I said, “Show me a Hong Kong spin.” He landed on his knee. His reactions were awful, his striking was poor, and he had no flavor. Lots of the other guys auditioning had similar problems. After auditioning them, I lined them up and told each one what they needed to work on, and I told “VT”,

“You have to work on everything.”

Dennis and I were the only ones they could afford to fly to India for the shoot. If the September recce was any indication of the talent level we were working with, then we were in for a very difficult shoot.

INDIA

Mumbai was a cool 85 degrees when we arrived in January. We went straight to the hotel and were instructed not to leave for the day because there was a riot between castes. The toilet was broken, and people kept trying to open my door at night. I expected some culture shock, but this was something else.

That said, Mumbai was an easier place to travel to than China, where I had worked 2 years earlier. In Beijing, you need a Chinese phone, Google is behind a firewall, and none of your social media accounts work. In India, your phone works, a wifi hotspot costs $10/month with 2GB/day, $3 gets you across town in an Uber, and the malls are brand new with fresh paint. You can buy anything as long as you know what it’s called. (I spent an hour asking for Q-tips in a department store. Turns out they’re called “ear buds”.)

Our shoot schedule: rehearse 1 week, shoot 1 week, repeat for 2 months. One day off per week on average. Our first tasks were:

1. Get the actors dialed in to the action.
2. Hire stuntmen.

We met the lead actors at the tiny stunt gym and asked them what they thought about the previz. Vasan told me,

“Let’s wait to show them the previz.”

The actors hadn’t seen the previz. They hadn’t even trained for it.

Dennis and I were rattled. Gulshan was still recovering from a torn ACL on his right leg. Abhimanyu was recovering from an injury, and Radhika had never touched a scarf like what we had designed, nor learned to use a chain whip or rope dart, skills she’d need if we wanted to deliver on the concept. The actors had trained to stay in shape, but they had no idea what kind of action we were shooting.

Then some stuntmen came to audition… the same stuntmen from the recce in September. The walls of the tiny gym we were in began closing in on us. 20 people in this tiny room made no room for training or rehearsal. We couldn’t even think.

I called it an early day and Dennis and I powwowed at the hotel. We freaked out. We ate at the café downstairs. Didn’t sleep. Jet lag settled in. Toilet is leaking. Wi-fi is spotty. No hot water.

In the morning I asked Vasan if we could move the rehearsal to a bigger space.

“Sure, sure. We can, we can.”

We met at a Himalayan tile distribution company where we had more space to think. Twenty stuntmen showed up on motorcycles to audition. We were told not to eat the buffet breakfast, so they brought us Subway sandwiches filled with chicken pot pie filling.

The actors show up and we meet with them one by one in the rehearsal area. They came in their workout clothes, but they wouldn’t need them. We just wanted to talk with them. Dennis and I did a mini-seminar with each actor, where we talked through the action design process.

“This is how we would shoot a front kick.
Here’s how we’d edit your scarf wrap attack.
When you’re on crutches, you can kick like this.”

”But if I can’t do that kick, can we change it?”

“Yes, we can change anything.
We can, we can.”

“How should I block?”

“See what’s comfortable for you.”

“Do we need to do all the action
in one master shot?”

“No, we’re not doing master shots.
Only the shots we need.” 

We filmed them doing basic attacks and exchanges. Their training with Prateek and Mhatung had provided them a solid foundation. We filmed them moving and reacting from different angles. We assured them that the action process would be catered to them because we were on their side, and we earned their trust, a critical first step.

Then it was afternoon. Dennis developed a quick basic audition for the dozens of stuntmen who had shown up:

• Shadow boxing – Slow, basic boxing strikes
• Kicks – Only the basic front, roundhouse, and side kicks
• Standing reactions
• Intercepted reactions
• Falls on an 8-inch mat

The qualified performers from the September recce had already proven themselves, but they went through the process as an example to the new guys. Most of the others had no boxing skills and wouldn’t perform the basic kicks we needed, couldn’t react, and fell without control. It would be a long process of teaching many of these stuntmen a critical lesson: to be a good stuntman, you have to be good at losing fights.

And then “VT” showed up. I gave another chance to see if he had listened last time. So I ran him through the process, but this time he aced it. “VT” had listened to my advice at the recce, stood in front of a mirror practicing, and came back with good strikes, great reactions, and painful falls. We hired him and held him up as an example of hard work and dedication.

We’d continue auditioning stuntmen between the shoot weeks. We ended up with 25 qualified stuntmen who could do the job. Lots of guys from the far East side of India – Assam, Nagaland, Nepal – would ace the audition. These guys were Jaika-level stuntmen before cameras rolled. Some had worked on indie action films in their home towns, like the Assamese Local Kung Fu, which is less a Bollywood film and more a Tony Jaa bash-em-up. They listened to American rock and watched Asian martial art films. Being geographically located closer to Thailand and Hong Kong meant these guys were infused with a totally different action language. A Tony Jaa look-alike from Assam spoke no English, but he understood how we were cutting on action, the need to cross the camera line depending on which shoulder the camera was over, and other complexities you only know if you live like a stuntman. They could do exactly what we needed and would help us lead the rest of the team.

LOCATION FIGHT PREP

We’re less than a week out from shooting the office fight. We need to get the actors up to speed, and they haven’t seen the previz. We haven’t placed the stuntmen in the scene. And we’ve only recently locked location and begun production design. We visit the set and position all the tables and props and breakaway tables, and draw it on paper so we can rehearse back at Himalaya Tile.

We test out the weapons we’re using in the fight. Art department has made some batons out of wood and metal. We say, “Make them softer.” They make floppy batons out of foam. I break one over my arm. “Make them harder.” We go back and forth like this for a while until we have the weapons that are right for the fight.

We rehearse the fight back Himalaya Tile for three days. First we break apart the previz we did in Oakland and fit pieces into the location, adding falls and props wherever we can. We mold the entire choreography around the location. We assign stuntmen to different roles based on skills. Vasan occasionally comes by and injects story notes, or we bug him for ideas. He casts an extra stuntman, a 6-foot-10 Indian basketball player, to add some flare, and we massage his character into the scene. The entire thing feels like we’re building a Jell-o mold, awkwardly holding it while adding pieces and pulling chunks out, ensuring it doesn’t fall apart before it gets to the table.

We tell the actors to rehearse slowly, as slowly as possible. They walk the choreography and request changes. Sometimes we let them change it, other times we push them outside their comfort zone. When they get it, they get pumped. More trust points. We do a blocking previz using my phone, our “design viz” or D-Viz. We decide we’ll figure out how to shoot it “on the day” (when we begin shooting) because that’s how we did it with Rope-A-Dope.

The stuntmen want to go at 110% speed during these D-Viz sessions. We assure them that they don’t need to show off. They’ve got the job already! We see all the other stuntmen circling us in the rehearsal space, watching intently, so we clear everyone out until it’s their turn to work. Suddenly there’s no pressure for them to perform in front of their buddies. It becomes an exercise for the stuntmen in moving slowly so they can retain everything in memory AND avoid hurting the actors. We audition some more stunt guys, change some casting for the roles, assure the actors that everything will be perfect, and we’re ready.

CAMERA

 

We clicked immediately with the Director of Photography Jay Patel. He spoke our cinematic language. We showed him examples of action we wanted to achieve and explained how dependent it is upon camera angles. He was on board with it all. His goal was to tell the most cinematic action scene possible, so we had a common goal.

We mapped out the set on a white board and did a rough shotlist, showing him our D-Viz and asking how we’d fit it into the location. We were very clear about something: we’d need a lot of shots, so the camera would need to be very mobile.

“Cool, cool.”

ACTION PROCESS – THE BEGINNING

The first day of shooting with a new crew is always slow. The cast has to get used to moving for camera. No amount of rehearsal can really prepare you for this. Abhimanyu’s first move out the gate was a 540 roundhouse. Not the easiest way to start your first starring role in a film… but he nailed it. All that extra training paid off. Gulshan was the same, having worked on his kicks and punches aided by a crutch. There are only so many kicks and punches you can do while on a crutch, and he had perfected them all.

With all this prep, we changed at least half of the D-Viz choreography on the day. This is totally expected. Stuntmen were shuffled about based on set design and story moments, which altered choreography. Camera placement changed, which also altered choreography. Vasan had brilliant story ideas, time ran short, and production design had moved a bookcase. All of that altered choreography too.

Clayton Barber producing the action in Rope A Dope 2. Barber told Jacobus, “A punch is just a punch. Focus on story, story, story.”

But the job of an action team isn’t to be dictatorial about the moves in the choreography. We’re tasked with telling an action story, and all departments are trying to tell that story. As long as the story was right, it didn’t matter if it was a right hook or a right roundhouse. As Clayton Barber told me,

A Punch is just a punch. Focus on Story, story, story.”

The number of people on set suddenly exploded beyond a hundred and we couldn’t see the actors 10 feet away from us. After “cut” I was yanking people away to make a path so I could give notes. The grip team were all standing about watching. Apparently hiring 30 “light men” is normal because of the low cost. Jay had commissioned them to build a complex, daylight-proof lighting setup. As a result, there was no need to move lighting during camera changes.

But when we changed the camera angle by 180 degrees, we had to move 100 people. Moving our 20-man crew on Rope-A-Dope for an axis change took three minutes, but each camera move in India took 15-20 minutes. On a shoot this big, despite having dynamic, deliberate camera angles, I’d need to make a detailed shotlist to save time. This way we could shoot everything facing West first, then reverse time and shoot everything facing East, etc.

This tactic saved us hours on the following days. Makeup, wardrobe, and continuity were all given the shotlist so they could plan accordingly. The crew had a rhythm and we were sprinting through the fight scene. When shooting Gulshan’s shining moment, we had a 2-on-1 exchange with falls and contact hits, with two over-the-shoulder shots alternating. To keep the energy high and avoid extraneous falls, we shot one way for segments 1, 3, and 5, and then the other way for segments 2, 4, and 6. The actors and stuntmen had to work through a mental puzzle, but it seemed to only engage them more.

This slideshow requires JavaScript.

Shooting the scene took the whole week. The scene with the weights came out beautifully and we worked in some new gags. The downstairs portion for the final 3 days involved the 3 heroes fighting 10 opponents in an open floor plan. So we sectioned off Supri’s action behind a large banner, Surya near a blown out window, and Gulshan in the center. This allowed us to move the camera freely without needing to fill the background with much action.

SAFETY

Safety in action films is handled a little differently overseas. I’ve been on shoots in Hong Kong and China where fake handguns are carried around on set like bananas. One guy snuck up on me in Hong Kong and pretended to shoot me in the back of the head, so I stripped the gun away and locked it in a safe.

Now I’m in India remembering there’s a gun coming later in the scene. I asked to inspect it. I was given a blank gun, the kind that fires shells packed with ½ or ¼ the normal amount of gunpowder, packed behind a paper wad. When you fire a blank gun at someone’s face, you can blind them, and if fired close enough the wad can enter a human skull and cause death. Blank guns kill.

So I asked the AD,

“What’s with the blank gun?
Who brought this?”

“Hold on, I’ll get him for you.”

I didn’t actually want to meet the guy. I wanted to know why there was a blank gun on set. Before I could fit the pieces together, an older guy introduces himself as the “firearms expert” and snatches the gun from my hands. While explaining how safe it was, he casually loads a blank into the magazine and chambers it. Dennis and I jump on him and strip the gun away, and he and his gun are kicked off set.

We held a safety meeting and created a safety process for anything that even looked like a firearm. Firearms would go through us for inspection, then be in the prop master’s possession until they were required for the shot, and then immediately back to the prop master. Any crew found breaking this rule would be fired.

 

One thing about India: they take processes seriously. Firearms on set were treated very differently after that.

Core stunt team for Man Who Feels No Pain. Left to right: Jacobus, Prateek Parma, Dennis Ruel, Arun Yadav

 

 

The office fight shoot was a huge success. The actors had performed incredibly well. And then it hit us: we hadn’t cast any stunt doubles yet… because we didn’t need any. We had totally forgotten about this, and the actors looked down at their hands and feet in amazement. We just hadn’t quite realized what we pulled off until then. I told Vasan,

The office fight shoot was a huge success. The actors had performed incredibly well. And then it hit us: we hadn’t cast any stunt doubles yet… because we didn’t need any. We had totally forgotten about this, and the actors looked down at their hands and feet in amazement. We just hadn’t quite realized what we pulled off until then. I told Vasan,

“That’s Rope-A-Dope style.”

Vasan laughed his out of control laugh. Nobody had done anything like this in Bollywood before, not on this scale.

We did hire a stuntman to do a backsweep. I tried to teach him, and he couldn’t do it. Another stunt performer was hired to take some hits to the arm, but the actor did a take anyway, and that’s the take we used. Our actors were able to do the action just fine.

THE BATTLE

After another week of rehearsal and auditioning more stuntmen we began shooting the final fight scene, the one that had taken up 22 minutes of our 45-minute Oakland previz video. We cut out at least half of it and, like before, re-tooled everything. New weapons were added and some performers had left, but we crafted another Jell-o mold for the massive finale.

I spotted some performers watching old Indian action films, with the lead flying around doing physics-defying maneuvers and stuntmen being yanked back on wires by the dozen. An issue was raised:

“I think we need to do some spectacular move.”

We reiterated that the action was grounded. It was comedic. These are people, not superheroes. They said,

“But it’s very common in India to perform superhuman feats.”

So I gave this one just to hold them off:

“We’re not making a common Indian action movie.”

Were we out of place? We seemed to be fighting for the team’s trust again. What would it take to convince them that we weren’t doing an Avengers film? That we were doing a Jackie Chan movie instead?

 

 

NOT JUST FALL GUYS

It came time for a stuntman to wreck hard on the ground. We had covered the ground in a thin, grey mat, but wanted to protect him more. I gave him direction for the fall. For any given fall, Dennis or I would throw ourselves onto the ground to demonstrate how to do it. They always paid attention after that. I asked this stuntman where his pads were. He shrugged. I asked if he had any pads. He said,

“I don’t need pads for this fall.”

I began to understand how it works in the Indian action industry. The action team brings in a few dozen fall guys for a scene. They randomly pick fall guys to take knocks. Padding seems to be optional. If they get hurt, they’re swapped out for other fall guys. Continuity doesn’t matter. They’re treated like interchangeable fall guys.

But Vasan had a different vision for Man Who Feels No Pain:

“In the 70s, there were so many Bollywood action movies where you remembered the bad guys’ faces. The stuntmen were actors too. Everybody had roles. That’s what I want to do.”

The stuntmen in our office fight weren’t interchangeable fall guys. We see their faces clearly. Some have conversations. Some make phone calls and have interactions with major actors in the scene. After they lose the office fight, their boss tells them they have to find the heroes and fight them again. In the finale, 80% of them come back, covered in bandages and splints, along with a few more fresh bodies. The stuntmen weren’t just fall guys. They were actors now.

So, when a stunt guy told me he didn’t have pads, I gave him pads and reiterated how important he was for the scene. There was a suddenly change of tone. Telling stuntmen they aren’t just random fall guys results in instant buy-in. These stunt guys were in it for the long haul.

They sprung for the pads. The first time, I asked one of the other Indian stunt guys for two knee pads and a hip. He delivered the request to another stunt guy, who told someone from production, who told a producer, who told someone from crew, who told a PA, who ran off behind a building. After ten minutes, the PA brought us a cup of coffee.

From then on I carried a backpack around with my own pads for whenever a stuntman needed them. Dennis or I would roll his pant leg up, slide it up onto his knee, ask him if he had any questions, demo the fall for him, and Jay would get the shot. We never wasted a single wreck.

WHAT MAKES A BREAKOUT

Midway through the finale, production brought in the rain machine. It was a monumental effort, but the look was incredible, leaving puddles of water everywhere, with the performers all fighting in water-soaked costumes.

Vasan wanted a fight using the Phantom camera, shot at 1000 frames per second to create a super-slow-motion effect. I’ve used the Phantom before, most recently on Heart of a Champion. It’s good for showing fist hitting a face. It’s not good for complex things. A complex WWE throw will be so slow that the workings of the move will be revealed, and the move loses its magic.

Everyone in Man Who Feels No Pain wanted to do a Hurricanrana (a WWE move) because that was the hot, new move. At regular speed, it’s fun. It’s what Black Widow does in The Avengers. But when slowed down to 1000 fps on the Phantom, you can see the stuntman’s hand help Black Widow stay balanced on his neck and guide her into the move. In slow motion, a Hurricanrana looks like a bad swing dance move. Dennis had designed the tone of the action for the film, and we cut these throws and anything like them from the choreography. This way we grounded the fighting in the reality of the story. But some of the team kept pressing for it:

“We need to do a Hurricanrana.
The Indian people will want this.”

So I said,

Eric Jacobus adamantly opposed any deviation from the action “tone” of Man Who Feels No Pain. “Jackie Chan doesn’t fly around on wires. Neither will these guys.”

“We’re not making an Indian movie!
These fights will be on YouTube someday.
If we do Hurricanranas,
and make stuntmen fly into trees,
the top YouTube comments will be brutal.
We have to make an Indian action movie
where the top comment is,
‘Wait a minute, is this an INDIAN movie?!”

I told them to think globally and get a map. Put a thumbtack here in Mumbai. Then put pins in Thailand, Indonesia, Beijing, Hong Kong, Seoul, Germany, LA, New York, Vancouver, Toronto, and every major city in the world. I said,

“That’s your competition.”

We had broken the mold already. Our stunt guys were acting, and our actors were doing their own stunts, all of them doing fight scenes that had never been done in India. We really were in Rope-A-Dope territory.

“Listen to us,
and nobody will laugh at these fight scenes.”

Nobody watched Indonesian action movies before The Raid, but the Wikipedia page for Indonesian action movies is a mile long. Why did Raid: Redemption put Indonesian action on the map? Because Gareth Evans, who had never made an action movie, and Yayan Ruhian, who had never done stunts, thought outside the Indonesian action movie box. The team behind Ong Bak had made action movies for 20 years that nobody saw, but Panna and Prachya thought outside the Thai box and put Thai action on the map. Same with South Korea when Jung Doo Hong did Shiri, Mad Max and the Australian new wave movement, Bruce Lee with Fist of Fury, and Melville and Delon in Le Samourai. Breakout genre successes do not follow cookie cutter models. They take risks and look outside the walls. Indian action has suffered from bad editing, unintentionally bad performances, and generic camerawork for years, but India has all the talent in the world necessary to make an Ong Bak or a Raid. All they needed was a structural change, and we were the crew doing that.

Slowly, the team was getting it. We were a small production, but the script and the action scenes, where there were no stunt doubles or wires or dumb rapid-fire editing, were bigger than anything any of us had ever expected. We were solidified as a unit then and there, because we all wanted to make that breakout action film.

SMOOTH FINISH

The rest of the shoot was very smooth. I had figured out a shot-listing process that allowed us to rip through the fight scenes without having to move the hundred-man crew too many times during the day.

Radhika’s opening fight at the car was stellar. We employed more Phantom shooting, but by then we had a philosophy behind it. Abhimanyu jumped over a car and kicked the Indian coordinator Arun, Radhika had mastered her scarf attacks, and Gulshan did an incredible 100-man kumite scene. Try to identify the Sikh in the final round.

I had written a clause in my contract that gave us a first pass at the edit for the fights. There was no guarantee they would use my edit, but it was critical to the vertical integration process. On the last day of shooting, Dennis and I had our bags with us. We ate some cake, went to the airport, thanked our driver and flew home, with a hard drive full of footage.

THE EDIT

As an action director, editing your own fights is a huge privilege. You get to cut around your mistakes, and your eye is set firmly on telling the best action story possible. You apply your expertise and craft the final vision of the action. I spent a month editing the four major fight scenes. We had never settled for a lackluster take during shooting, so the edit was easy. It confirmed the entire process.

Then I laid a sound design with thwacks and whooshes to compliment the style of the fights. I told production it was critical to have the right sound design for action. Too soft a sound would undermine the action, and overdoing it would cause a separation. A balance had to be found.

I delivered one fight per week on average, and received very little feedback except,

“Cool, cool.”

I delivered the final OMF and EDL so they could put it in their timeline. Everything was in on time.

Then it was radio silence. The Indian team was swamped with post-production; editing the rest of the film, sound mixing, music composition, vfx to remove thousands of frames of Gulshan’s green-screened leg and erasing thin mats from hundreds of shots.

My job was done. I had to wait and see. There was no guarantee they would use my edit. I kept telling myself, “Just have faith, and wait and see, Eric.”

TIFF PREMIERE

Then a trailer dropped. Man Who Feels No Pain would premiere at the Toronto International Film Festival’s Midnight Madness. The trailer was hilarious.

Dennis and I flew to Toronto. Vasan got us and some friends some tickets to the show. We met Vasan, Ankur, Abhimanyu, and Radhika there. Vasan tackled me with a huge hug. He had been locked away editing Man Who Feels No Pain for months, dealing with reshoots and everything else.

The lights went down and the film showed to a packed auditorium, and the audience was there to have a good time. And they did. As a film, Man Who Feels No Pain starts off with an amazing Act 1 centered entirely on Surya’s youth. At about 30 minutes, we see the first fight, the Kumite. They used my edit and sound design.

This confirmed everything about the Vertical Integration process. I breathed a sigh of relief. The fights were exactly how they needed to be. They were beautifully performed by an incredible cast, shot by an awesome DP, and edited (and sound designed) by the guys who directed the action. The audience loved the entire movie. The action was just there to compliment an incredible story, the very story that sold me in the first place.

VERTICAL INTEGRATION RECAP

1. Script – We coordinated with Vasan Bala, who also wrote the screenplay, in order to modify any elements and themes throughout the script to maintain consistency with the action. If edits couldn’t be made to the script, we changed the action.
2. Pre-visualization (Previz) – Our team delivered varying levels of previz depending on the requirements of the production. Our first pass was primarily for the sake of reference (D-Viz) and when we began rehearsing we shot a more concrete previz with camera and editing suggestions (Pre-viz) followed by any last-minute modifications that had to be made due to the nature of the location or set (Set-viz). An action team’s agility is critical so that we can adapt and create whatever previz is necessary for the situation.
3. Choreography – The choreography was developed at various times: during D-Viz when we were doing character-based choreography, after auditions had been finalized to take advantage of the performers’ abilities (Pre-viz), and finally after sets or locations had been finalized to take advantage of the environment and ensure camera placement is still feasible (Set-viz). This final stage of choreography during Set-viz is when the blocking choreography is solidified. Then, when cameras finally rolled, we’d change half of it anyway, and we knew this would happen, so we never belabored particular moves or character blocking patterns.
4. Camera – We were blessed to work with Jay Patel who helped us craft a very precise shooting style for the action that was consistent with his vision while maintaining continuity of action. The film was shot fairly traditionally with no shaky cam and a little bit of handheld, so we worked this into our action as well. The framing would always cater to the performance of the actors, the way a Bollywood dance scene would be shot. If the shot was lacking, we’d rarely compromise with a camera angle change, but re-target the choreography to the camera. Most of the shooting of action was piecemeal, meaning the shot has a specific place in the edit. Only 5-10% of the action shooting was coverage-based, meaning we shot the same choreography from a couple angles, which was usually to simulate a “dialog” between characters fighting. 40% of the time we had a second camera running, but I only used these “extra” angles for 2-5% of the edit, usually only for covering dramatic pieces in the scene.
5. Performance – Vasan’s vision was to cast actors who could do stunts, and stuntmen who could act. Our team held 2-3 audition sessions per week to cast stunt performers based on their physicality and reactions, auditioning over 100 performers in the end. Vasan would then judge them based on their acting abilities. For acting-intensive parts, Vasan would cast first, and we’d judge their physicality. With all this considered, our team would recommend a role for them in the choreography. If the role was already determined, we would modify the choreography to fit their abilities. The principle actors had already been cast and were physically gifted, so it was a matter of training them and getting them comfortable with choreography, making changes as needed as long as they were consistent with the story. We would usually lean on Vasan whenever making decisions like these.
6. Editing – My contract stipulated that I could do a first-pass edit of the fight scenes and make best-take suggestions, and they sent me home with a hard drive containing all the action footage which I would edit in Premiere. There was no guarantee they would use this edit, nor should there ever be such a guarantee. But since the director was instrumental in the piecemeal shooting, he was already envisioning the edit, so the final edit didn’t differ much from his expectations. There were some changes to pacing for less linear parts of action scenes (less than 5% of the total action edit), but our best takes were employed. For every minute of action I spent about 5 hours editing, so for 2 small fights (2-4 minutes each) and 2 large ones (12-25 minutes each) it took about 5 weeks after sound design.
7. Sound Design – I was not obligated to deliver a sound design, but I did one anyway to demonstrate the vertical integration of action and sound. Without weak sound effects, a fight can become flat, or if the effects are too exaggerated the fight can balloon into comic proportions. Sound effect timing is also critical to action continuity. Some accidental misses can be made to sound like hits with slight changes in the sound timing, and weaker performances can subtly brought up to par with smart sound design. They employed the sound design in the end. I delivered everything to production in EDL and OMF format.

FINAL THOUGHTS

In Man Who Feels No Pain, we were hired to deliver a package we call vertical integration, wherein all departments facilitate the action vision.

But most importantly, Vasan and RSVP graciously gave us the opportunity to implement our process for creating action for Man Who Feels No Pain: an end-to-end, vertically integrated action model, which was both agile and technologically robust. It was the latest iteration of the model we used in Rope-A-Dope and Blindsided: The Game. We couldn’t test this model in a lab, so Vasan and RSVP gave us the chance to use it on a real project. We thank them, the cast, and the crew greatly.

Enjoy Man Who Feels No Pain. You’ll love the story and the action. That’s as good as it gets.