Go ahead. Read it. I’ll wait.
Done? Good. Its’ wrong. Wrong wrong wrong. If you are a filmmaker, low-budget or otherwise, don’t even start thinking in the terms suggested by that article/post.
So why am I picking on some blogger? Well firstly, Farhan has published a book and seems to feel authoritative on the subject. Second, the VFX Society is retweeting it. But we in the VFX industry are already under assault. And it’s time we started taking this all more seriously. That got me thinking.
I went to NYU Film School at Tisch, back at the turn of the century (1996-2000). Back then, there was no real curriculum for filmmakers on how to use VFX. I did actually learn Power Animator and Maya by sneaking into a 3D animation class that I shouldn’t really have qualified for. But filmmakers should not need to learn 3D animation tools in order to work with VFX.
Directors and producers don’t learn architecture in order to get sets built. I don’t think anyone is teaching low/no budget filmmakers how to approach low/no budget VFX. And further, I don’t think anyone is teaching how it’s best done professionally either. Not from the standpoint of the filmmaker anyway. There’s a ton of people willing to teach you about every piece of software on the planet, which we use to make VFX. Unfortunately there is very little scholarly or even practical material out there, on how best to direct and manage the process from the production side of things.
So here we go. This is what you actually need to know. I will borrow some structure from the afore mentioned article/post, to best allow comparison and contrast.
Find out who you need to do the job
Farhan suggests getting familiar with various VFX vocations. Animators and modelers and effects animators… oh my! He asks, “Do you need a separate compositor?” Wait what? Compositor? What’s that again?
Stop. Stop right there. This is not a making-of-featurette where the studio lies to you about how cool and complex filmmaking is. Wrong approach.
Every producer knows, you start by breaking down the script. It doesn’t matter if your are a one-person show or a mega-production. Once you break it down, you assign all the things in your script that need to be some form of special or visual effect, to that department. How do you know if it’s a special or visual effect? Probably because it’s in the script, and you need to get it in front of the lens, and you have no idea how to get it on set in a reasonable way.
Now, who do you need? Well, really you need a Visual Effects Supervisor (VFX-Sup). That’s someone who’s going to take responsibility of handling the VFX from a creative and technical standpoint. You might also want a specialized Visual Effects Producer (VFX-Prod). But wait… this may be a low/no budget production. Let’s stop before we start piling on more production staff, and take a closer look.
Basically, if you think you can beg/borrow time and get someone to be your VFX-Sup, you should. They will be much better equipped to make suggestions and decisions. The fact of the matter is: If you need a Production Designer, you’re going to get one. You are not going to try to learn to do production-design yourself. Not unless you happen to also be a Production Designer. And even then, it’s usually not a good idea to do it yourself if you don’t have time. VFX are no different. They need their own department head.
You might look to professional VFX Artists and offer them the Sup job on your project. Getting time with a production on-set should be a big draw for up and coming VFX Artists.
I do not recommend making a deal with a VFX Vendor in which their own VFX-Sup will be your VFX-Sup. It is most important that your VFX-Sup both be aligned with your Director creatively, and also with your production monetarily and career-wise. VFX Vendors will ultimately end up in a conflict of interests with the production. They are a vendor. They are not you. It doesn’t need to be adversarial. But it does need to be clear and respectful. Muddying the waters is not a good idea. Then, you are always able to negotiate clearly.
One way to do without a VFX-Sup is to see if you can keep shots completely separate from one another, such that you can contract them out as whole-shots to individual Artists. Cities like New York often have a freelance artist-base that consists of more of what we call “Generalists” than “Specialists.” A good Generalist can often take a whole shot from start to finish on their own. In some cases, they may be able to subcontract some tasks out themselves. Los Angeles also has “Generalists” but they’re mixed in with the “Specialists” by the nature of the market. So you’ll need to clarify what you’re looking for in those kinds of markets.
Keeping shots separate from one another means avoiding “interdependencies.” By this, we mean “shared-assets.” If multiple shots have the same CGI character in them, they can’t really be treated separately. You can tell this very easily from your breakdown. However, if you have just a bunch of wire-removals or rig-removals, those shots don’t require continuity. They just need to each be done well. A quick set-extension sequence might be okay if you don’t see the same extension twice.
You might be able to bend the rules a little bit if the shots are simple. You might be able to send out three shots that require the same set extension, to the same artist. But you want to be realistic about how much time you have with that artist. If you’re giving them 15 shots but you only get back 5, you’re going to have trouble giving the remaining 10 to someone else, if you can’t get them to match up to the 5 you already have.
What you should not do, is try to build your own team of “Specialists” and try to have them collaborate. You don’t want to be managing shared assets and giving them out to different artists at different times. You don’t want to manage that kind of digital-asset dependency system.
Were you to do that, you would then be taking responsibility for shared assets and production-pipeline, in an area that is one of the most complicated production processes in the world.
Some day, out-of-the-box tools will handle this all so well, it will be like photoshop and illustrator. But today, the tools do not work so well that you can just get them, and throw people together.
Have you ever seen an Illustrator Artist and a Photoshop Artist complain about each other’s layers and organization? Imagine that but 1000 times worse, with over 100 separate assets and files to complain about and work with.
If you can’t keep your shots separate, you probably need a VFX-Sup. Or you need to become a VFX-Sup. But that could take a decade. So do what you’d do for a production designer. Get creative about finding the right person to do the job.
Whatever you do, don’t worry about compositors and lighters… and which should really be doing the final light.
If you have complex asset dependencies, your VFX-Sup will likely either seek a Vendor, or build their own pipeline-standards for a collection of specialists. But either way, this should all be normal to you even in a no/low budget world. Getting it done on budget, cheaply or freely is about negotiation. It’s rarely actually about doing it yourself.
Also, consider that in an extreme low/no budget situation, you’re not likely to be able to attract the cream of the crop. Work that comes back, may be useless. But, if you aren’t paying for it, you don’t need to cut it in. It may make sense in some situations; to commission work early and assume you may need to send it out to a completely different artist a second or even third time. If that’s your strategy, build it into your schedule. Time is money after all.
Define and gather reference NO! Design!
Farhan suggests gathering reference. Actually, that’s too narrow. What you want to do here is design.
Design can include reference. But the point is to lead. You want the city to look a certain way? Gather relevant reference. But also work with your VFX-Sup and Production Designer to define it further. Design your movie. You want a CGI character? Sculpt it. Or digitally sculpt it. Do iterations if you need to.
You can actually get pretty far along building production assets here. Tools like zBrush are doing a phenomenal job of allowing artists and sculptors work with filmmakers in a very fast and iterative manner, for very little money.
The important thing to do here, is take the time to do the creative work that’s going to inform your further planning and sequence designing. Every dollar you spend here is probably worth 10 further down the line. James Cameron worked in pre-production on Avatar, for years. What was he doing? Designing.
Previs No! Sequence Design!
I used to be a Previsualization Artist. I worked at Pixel Liberation Front as an Artist, and eventually as the Staff Technical Director. I know my Previs.
And most importantly, I know when not to do Previs. Too many projects jump into Previs without understanding why they are doing Previs.
Previs is done for two reasons usually. Either for creative development, or for technical planning. Often for both.
But at its core, the purpose of Previs is to get you what you need, in order to get through principal photography. So if you really need to Previs a sequence for creative purposes, and you can’t shoot without figuring out those creative problems, go for it. But make sure you’re doing it for the right reasons.
Usually, you’re doing Previs because you need to understand exactly how to accomplish a shot or sequence. Or you may need a visualization of it, in order to get your crew to understand how to execute it.
Once you recognize that you need to do technical planning, you may realize you need to do the full creative development on the sequence. You may need to do this, in order to lock down the parameters you’ll need on set. Such as the lens and camera position, or the size and position of a green-screen. Sometimes, it can be as precise as outputting an animated camera for a motion-control rig. Though that’s unlikely in a low/no budget scenario.
Ultimately, Previs is about committing to your sequence. And that’s the hardest part to do. But if you’re going to spend your budget efficiently and not go over-budget in post-production, you need to be diligent about your Previs and commit to it. Take it seriously.
To even start Previs, you usually need to start with an idea of what your sequence is. Then you need to do your best to map it out on paper. Storyboards are best. But even a simple shot-list will do.
It is possible to hand a screenplay over to a Previs Artists and tell them to “go!” but it’s not the best way to work. There is still a lot of tedious digital asset development that goes on during a Previs project. And any time you can save by giving them a more refined and directed idea of what the shots are, the more time and money you will save.
That being said, Previs is the time to experiment. Once you’ve got something built, don’t be afraid to change it. Yea, they might groan. They’ll get over it. There is a reason why experienced Previs Vendors charge by the day. It’s best for everyone. It’s also best to budget the right amount of time for experimentation.
When I train VFX artists, I want them to fall in love with their shots. When I train Previs artists, I train them to be able to be dispassionate filmmakers. Sometimes, you can Previs an entire sequence, and then have the sequence cut from the film before it’s even shot. That is not a failure. That is a success. Previs Artists are there to help you go through creative development and sequence development. And sometimes, cutting out shots and sequences is best for the film.
But I’ll stress again: The most important part of Previs, is that you keep your head about you. Why am I doing this Previs? What do I need to get done here so I can shoot? Am I moving toward getting what I need? Can I commit to these decisions?
Visual Target No! Postvis! More design!
After principal photography, you need to get your rough-cut put together. A rough-cut is effectively a re-write. So you can’t really delve out VFX shots to be worked on, until you’ve finished that re-write.
But there is a chicken-and-egg problem here. You may not actually be able to cut together your film without some kind of representation of your final VFX in the shots. Some elements may be so completely disassociated as to make no sense without some VFX done to them.
This is where Postvis comes in. It’s often similar to Previs. But it’s disimilar in that you are using rough versions of final elements (plates) to slap together enough of a representation of the effect, to allow the editing process to continue.
Postvis effectively un-blocks the creative process here. And so you are almost ready to hand over your plates to your VFX-Sup or VFX Artists. But you need to stop here, and make sure every last bit of your design work is done. Why you ask?
Because VFX is expensive to change. It’s not the easily malleable putty that we like to think it is. Sure, it’s often much more malleable than footage alone. But it costs money to rebuild and rebuild at that kind of quality. You want to shoot for no revisions. You want to provide every piece of design, previs, postvis and reference such that the VFX Artist has no choice but to make exactly the shot you’re looking for here.
If you don’t think that’s where you are, you probably need to do more design. Possibly more refined previs or postvis.
But finally, once you’re good, you can do a “turn-over” to your VFX team.
Integrate Into the Backplate No! Supervise, Iterate and Approve.
Now, I did say that you want to minimize rebuilding and revising. And that’s true. But that doesn’t mean it’s not part of the process. You should want to see work products early and often. You will want the VFX-Sup to hold your hand a bit and tell you what you should be looking for in early work-products. If it’s unclear what part of an image is temp, versus what part is actually being evaluated, ask. Don’t be embarrassed. This is the most complex production work on the planet. And it’s constantly changing.
The more work you look at, the less likely it’s going to go in a direction you don’t like.
Try to avoid “noodling” which means, making notes and comments for the sake of making them. Try to balance your desire to say something, versus your desire to move forward. If it’s important, say it, for sure. But make sure the project doesn’t get hung up on comments and notes that aren’t really making it be what it needs to be.
Be comfortable asking what the next steps on a shot might be, so you can judge best when to bring things up.
Try to provide further reference. Consider if you need to put something on hold, in order to do further design work elsewhere.
Keep a target schedule for shot approvals. Work with your VFX-Sup to make sure you are on track to get the shots done on schedule.
And whatever you do, don’t ignore the process, and then complain that the shots are not what you wanted, during the color-timing process. That kind of behavior is beyond unprofessional. Some day, producers and even studios will understand that well enough to punish filmmakers who don’t embrace the VFX process as their own process. When that day comes, filmmakers who truly get it, will prosper.