The VFX Factor: Eye-Opening Learning Experiences with Jeff Kyle

Jeff Kyle is a flame artist at New York headquartered post production studio Assembly. A seasoned post-production enthusiast, with an insatiable drive to learn new software and concepts, Jeff is often pushing the boundaries of what is creatively possible with the technology at his fingertips. He recently graded a Sundance award-winning film, Kokomo City, using flame software, a tool traditionally used only for compositing. 

LBB> What’s the biggest misconception people have about VFX?

Jeff> If I had to pick one, I think the biggest misconception is just how creative the problem-solving process can be on every shot, not just the stand-out ones. As it stands, we don’t run the shot through the “screen-comp node” or the “fancy explosion node” or even the “paint out the boom mic node” and then let the computer do its thing. It’s a fluid process with a number of steps and a lot of room for variation. How we get from start to finish will depend on the artist’s experience, how much time we have, and what we understand to be the confines of the shot. I think the creativity is a lot more subtle on some of the more mundane shots, but it’s certainly still there if you go looking for it, and to me that is one of the more satisfying parts of the job.

LBB> There are two ends to the VFX spectrum - the invisible post and the big, glossy 'VFX heavy' shots. What are the challenges that come with each of those? 

Jeff> I’ve spent a lot of time with invisible post shots, and I think one of the biggest challenges is just how much of a mixed bag it can be in terms of what needs to go into the work and how to estimate how long it will take. I’ve had groups of shots where one will take three minutes and one will take three days. A lot of the time we’re involved early on and we’re asked to bid on how long we expect certain tasks to take based on storyboards or just concepts. Of course, we build into our bids some stipulations that protect us so we aren’t held hostage to some of the guesswork, but it can be very challenging to try to account for the things that are not yet known. This is of course in contrast to the ‘VFX heavy’ shots where good planning is the difference between having a miserable time and having a great time. Of course, that’s the case with everything, but when it comes to those big shots, it’s all the more important. I don’t have a lot of repetitions in for those big shots, so it’s hard to speak too definitively about them with any authority. 

LBB> As a VFX person, what should directors be aware of to make sure you do the best possible job for them? 

Jeff> I’m sure this is said a lot, and I’m sure it always depends on the project and the budget, but getting involved as early as possible is always incredibly welcome and useful for getting ahead of a number of problems. Getting another pair of eyes on anything can be helpful, but getting someone with a VFX eye in particular to take a look on the front end can save the day down the road. Plus, with some of the new technology out there, we’re now not only able to view the camera feed remotely, but we can also gain access to the footage almost immediately after it’s shot. With all of that at our fingertips, it’s hard to pass up the chance to get involved if we can do so from the comfort of our own work-from-home setup. 

LBB> VFX is a true craft in the classic sense of the word. Where did you learn your craft?

Jeff>I started my VFX journey in high school, learning after effects from a book and from the legendary Andrew Kramer tutorials on YouTube. As I delved deeper, I subscribed to sites like Lynda and Digital Tutors, absorbing as much as I could about the rest of the Adobe suite and Cinema 4D. But it wasn’t until I landed my first job in the industry as a production assistant that I discovered the programme I’ve been using most of my career - flame. While there are a number of tutorials for flame, and I was no stranger to tutorials, I often walked away from them with more questions than answers. Luckily, the flame artists I was assisting are great people and spent an insurmountable number of hours teaching me everything I needed to know throughout my time as an assistant. I owe everything to my mentors at Nice Shoes who gave me the time of day and helped shape me into the artist I am today. Once the training wheels were off and I was running jobs myself, I took to the larger flame community to contribute what I knew and also learn from the hundreds of other artists out there willing to share their knowledge. All of that, combined with being a (relatively) active member of the Flame Beta programme to stay sharp on the latest upcoming features, brings me to where I am today. 

LBB> Think about the very, very start of a project. What is your process for that? Do you have a similar starting point for all projects?

Jeff> When it comes to your average job, I’d say the start typically involves some kind of kick-off meeting with my producer and any necessary parties such as the colorist, other artists, or other involved departments to set expectations and understand what’s going on. But after that, it’s a little hard to generalise about the next steps, since I work on such a wide variety of jobs with different needs. Some are simple, quick turnover jobs that really require very little thought or preparation, while others are quite large and might span multiple months. When it’s the latter, I might be invited to take part in multiple meetings surrounding planning and occasionally I’m invited to the shoot to ensure everything goes smoothly and nothing is missed. For most of my career I’ve been involved with commercial finishing, but the last few months I’ve begun to explore the world of episodic shot-based work which is quite a different beast. The kick-off is the same but many aspects of the work are quite a bit different and has been an eye-opening learning experience to say the least.

LBB> We imagine that one of the trickiest things with VFX is, time issues aside, deciding when a project is finished! How do you navigate that?

Jeff> On the one hand, there’s something very satisfying about cleanup work because it’s binary - remove that wire; take out that logo. It’s satisfyingly easy to know when that work is done because the goals are so clear. It was there before and now it’s gone; job done. The next level of work in mind are things like screen comps. That’s almost completely binary because the screen is either in or not, but there is still certainly some grey area with the level of polish on the finished comp. There are a lot of little things you can do to bring it to the next level. Getting the track to be a little better, adding the perfect amount of motion blur; a laundry list of things where if you have the time and the patience, you can take it from a great shot to a fantastic shot. Experience allows you to make those decisions about what to do and when to do them based on the needs of the shot and the expectations of the project. And then finally there are your real creative shots. Heavy compositing with what feels like an infinite number of levels of polish that can be added, time and budget permitting. I’m the first to admit that I haven’t spent a whole lot of time on those shots, but I can imagine that for those artists who work on those kinds of projects, knowing when the shot is finished is quite the challenge. 

LBB> Is there a piece of technology or software that's particularly exciting you in VFX? Why?

Jeff> It’s been growing for a few years now, but the machine learning advancements are super exciting to me and I’m having an absolute blast with it. We have a few integrated machine learning tools inside of flame that have some niche uses I’ve been able to take advantage of on a handful of projects now. There’s face detection and sky detection for generating mattes that are great for colour correction. It might not seem like much, but for anyone who has tried to isolate those elements when there is a lot of colour variation in those areas will know that it might not be easy. The machine learning isolation tools don’t rely on similar colours - they just know what a sky is and what a face is and they can get you your matte. But another company that specialises in machine learning named RunwayML took it to the next level, using a wonderfully simple web interface to allow the user to click in a few places on the image to generate mattes. It’s not as good as full-fledged roto made by a real person, but sometimes the machine learning roto will be good enough for the needs of the shot, all for five minutes of your time. The final noteworthy machine learning advancement is that someone in the flame community found a GitHub repository of a machine learning time warp tool and figured out how to make an interface for it inside of flame using Python, and the whole flame world has been forever changed for the better. The results it produces are life-changing, and it has become a must-have tool for me on every project that needs a timewarp. I can’t wait to see where machine learning continues to improve, and what’s next for the tools that make our lives a lot better. 

LBB> And as real time tech and games engines become ever faster and more sophisticated, how do you see that shaping or changing the role of VFX and its place in the production pipeline (e.g. thinking about things like virtual production)?

Jeff> I’m in awe of productions like ‘The Mandalorian’ using virtual sets that can be manipulated live. I’m sure some might be worried about virtual production reducing the need for VFX since it can and does replace the need for certain things like green/blue screen compositing. I think that’s probably the wrong way to look at it. It allows you to see things in context like never before - very close to being on location. But it comes at the cost of not having the flexibility to change it later unless you go back to the green/blue screen approach. Even when going all-in on virtual production, there will always be more things to do, adjustments to be made, and issues to be fixed. This is just another tool in the toolbox and I think it’s a net gain for the industry as a whole with what it brings to the table. 

LBB> VFX is a craft that relies on you really looking at nature - how light works, how gravity works, the mannerisms of a kind of creature, how crowds work, skeletons, explosions… whether it’s animation or compositing or anything else… So how do you like to approach the research side of your job? What’s the most random or intriguing thing you’ve learned from working on a project?

Jeff> I spend less time with creatures and explosions and more time with some of the more subtle natural elements like light and shadows. Observing how those two natural elements interact in person is quite an accessible thing to do because they’re all around us, so taking a good hard look at those real-life examples is quite the accessible exercise. But once you’ve computed enough lighting effects and shadows, I think you tend to get a feel for some of the tricks to get it over the line and you don’t necessarily need to start from the research stage again. Alternatively, when I was still learning about and working with all things 3D, that’s when research was a lot more relevant. I was working on a project that involved creating CG columns and I stumbled upon an architectural ornament called a volute, which is a scroll-like element. This was about 10 years ago so it’s a little hazy, but what I do remember is getting absolutely obsessed with how they’re made and where they’re used. It’s a little tricky to model, and it seems pretty obscure, but once you know what it is and what it looks like, you start to see them everywhere (especially on buildings in Manhattan where I work). Being in the VFX world, and more specifically the world of 3D, really shapes the way you look at pretty much everything. Since we’re in the business of creating, modelling, lighting, enhancing, distorting; that lens really has a big impact on what we see and how we see it, and I think that’s pretty cool.

LBB> When you’re watching a VFX-heavy ad or movie, what are the tells that you look for to figure out how well-crafted it is?

Jeff> There sure is a lot to look out for with VFX-heavy media. The biggest standout attributes to me are the little things. The first thing that comes to mind is a scene in ‘Avatar 2’ where we submerge underwater for the first time. The underwater world felt so incredibly alive–not because there was one or two impressive CG fish, but rather a whole ecosystem of big fish, small fish, medium fish in the foreground, midground, and background; a smattering of underwater plants moving with the water; more bubbles than I could count; and a handful of CG people swimming around. Obviously, ‘Avatar 2’ is at the top end of VFX-heavy movies and is the cream of the crop, but I’m quite interested in secondary and tertiary additions to make things feel alive and quite well crafted. 

LBB> How did you first get into the industry? What was your very first job in the industry and what were the biggest lessons that you learned at that time? 

Jeff> After years of being interested in the industry, it was toward the beginning and middle of college that I finally decided I wanted to pursue a career in post-production. When I finished college, I looked wherever I could and found an internship at a tiny post production studio of about four people total where I was doing odd jobs around the office - refilling the coffee machine, going to get groceries, and running hard drives around the city. In my downtime I sat with the After Effects artist and learned about what it meant to work on real projects. Since an unpaid internship wasn’t necessarily sustainable for a recent college graduate, I started looking for a full-time job in the industry. I applied all over the place and lucked out with a series of interviews for a production assistant role at a mid-sized commercial post-production studio Nice Shoes. I was incredibly fortunate because I was told there were a ridiculous number of applicants for the role, but the person doing the hiring said that my standout feature was the fact that I had delivered hard drives around the city at my previous internship and the job would involve a lot of that. I guess sometimes it’s little details like that that make all the difference, because getting that job was an enormously important moment in my career. 

But it was in that role where I learned just how true it is that you make your own luck. The day-to-day of a production assistant can be a little mundane: you’re picking up food orders for clients, helping engineering with their projects, managing shipping and organising hard drives that come in and go out, and helping out with various office tasks and events. It’s easy to get a little disillusioned by these mundane tasks, but I learned early on that regardless of what you’re doing, how you handle it acts as a great representation of who you are as a person, and ultimately gives other people a good idea.

While the day-to-day tasks of the job were simple and mundane–picking up food orders for clients, helping engineering with their projects, managing shipping and organising hard drives that come in and go out, and helping out with various office tasks and events - I learned the important lesson that you make your own luck. Regardless of the task at hand, it's important to approach it with enthusiasm, professionalism, and most of all a positive attitude. I realised that how I handled these seemingly small tasks was a reflection of who I am as a person, and could and does have an impact on my future opportunities within the company. It was through this I learned the importance of making the most of every opportunity, no matter how seemingly insignificant.

LBB> What was your first creative milestone in the industry – the project you worked on that you were super proud of? 

Jeff> One of the biggest creative milestones in my career was when I landed my first colour grading job as a flame artist. After exploring colour grading workflows and learning the craft for several months, I was given the opportunity to work on a 10-minute short film. This project was particularly special to me because it was my first real grading experience in front of a client and it also included a handful of comp shots which made it feel like it was the perfect flame experience. Being able to offer both services and deliver high-quality results quickly was a great feeling, and it set me on the path I continue to follow today in pursuing flame colour jobs. I'm really proud of how that project turned out, and it's a reminder of how far I've come and how much I've learned since then.

LBB> From a VFX perspective, which ads have you seen recently that you've been particularly fond of and why? 

Jeff> The Super Bowl ads are still fresh in my mind, and one that really stood out to me was Snapchat's commercial on augmented reality. It was both mesmerising and disturbing, which made it all the more memorable. I’m a sucker for transitions, and many of theirs were buttery smooth. The face effects were over-the-top and almost scary, but very on-brand for them. Overall, it looked like a really fun project to work on.