VFX: Is now the time to rethink this billion-dollar industry?
Currently in downtown LA, in a renovated top floor, there is a movement to discuss and perhaps evolve film VFX production, in other words to re-invent the business model of how film’s visual effects are made. This is not a side discussion, it has started at studio level, with Oscar-winning visual effects supervisors and senior studio production executives. It is not a fight or a protest, it is simply a business based discussion that could be framed as “there has to be a better way”… and answered with a model transposed from outside the film industry. It is as simple as it is radical, but it requires a whole new approach to filmmaking much more reflective approach one might call ‘visual development’.
The problems facing the visual effects industry are serious and complex, as seen by the number of companies failing on razor thin margins to maintain their businesses and are either closing, or consolidating. From extremely high profile companies such as Rhythm & Hues at the very point of winning a VFX Oscar, or Digital Domain restructuring, to companies that are less well known such as Look Effects closing. Between 2003 and 2013, 21 visual effects companies closed or filed for bankruptcy, yet of the 50 highest grossing films of all time, 49 are what one might call ‘effects films’.
The model is being advocated by Ben Grossmann and the team at MAGNOPUS, but they are quick to point out this is not a solution that is centered around their company alone. They would like to advocate the change, but they hope other companies will adopt, embrace and use the new approach. They may be the advocates, but VFX is a collaborative sport and they seek broad discussion and involvement.
What is the new streamlined and potentially long term business model? It owes much to manufacturing production outside the VFX industry, to understand it one needs to summarize the problems facing the industry.
Inventing the future
While companies do large amounts of visual effects work that is repetitive and well understood, there is a significant body of work that involves inventing new techniques and processes. One hallmark of visual effects is the notion of producing something no one has seen before. While effects can be deployed just based on cost savings (for example removing signs digitally in a street, instead of manually during filming), VFX are often used in spectacle. The effects may be invisible or extremely obvious to the shot, but many award winning effects are classified as such due to their originality. While most companies prepare quotes as accurately as possible, it is often hard to bid a look that has yet to have been developed. How long will it take to produce a new creature that is ‘unlike anything anyone has seen’? How long will it take to make an ocean react in a new and magical way to that creature? This research component can be reframed as a risk component, given that the visual effects company will unlikely be given an explicit R&D budget, it must factor in a component of research costs to the overall bid. In a real sense the VFX company assumes that on a bid of size X, there will be enough value in the bid to cover this unknown amount of research and development.
While this approach could work with healthy margins, the competition in visual effects is so great that it is unlikely there is much scope for padding the bid. There is also another unexpected side effect from the producer’s point of view. Unlike other departments such as set construction or wardrobe, the VFX bid becomes an opaque large number. It is not broken down on hours and materials. The producer must accept the bid without fully understanding the internal numbers that generated the master figure. They are not invited to see a line by line costing, since the VFX house believes R&D along with many other ‘normal’ non-shot related costs would be rejected by a client. These other costs include archiving, system admin functions, networking costs and training, to name a few.
What is Visual Effects anyway?
Following on from the point above, a producer is both unable to see the line by line item costs of a very large part of their budget. They also assume perhaps that the visual effects are the digital equivalent of the previous physical effects, but now done in the computer. The problem with this view is that the visual effects budget covers much more than any special effects budget ever did.
If we assume a period action cop drama, set in some early part of the 20th century in an urban environment as a mythical example, then what is effects and what is actually modern filmmaking?
For example, if there is a complex scene where there is a gun fight for example, the line item of Visual Effects might cover:
– gun flashes – replacing traditional physical effects for reasons of safety
but also it might include:
– removing modern signs, set extensions etc which might have been art department previously.
…and this is not an isolated example.
It might cover a wide shot of the city, replacing traditional glass matte painting effects
…but it might also include:
– wardrobe fixes (costume)
– fly away hairs on the leading lady and digital tattoo removal (hair and makeup)
– adjustment of street lighting (lighting department)
– camera shake removal and slight reframing (camera department)
– combining of two takes into one (editorial)
– falling shot dead guys (stunt department)
– crowds scattering in the background (casting, 1st AD, costume, etc)
– sky replacement (color and lab timing)
– etc etc
In the one VFX category which will run to tens of millions of dollars on a Hollywood film there are actually multiple departments being serviced, of course one of them is effects, but the move to increased budget size has not lead to greater control and influence by the visual effects team. It has lead to less understanding of the financial breakdowns of modern feature film production. In an absence of knowledge, facing an opaque VFX bid, the producer can only seek to lower the overall bid. If VFX represents a vast percentage of the budget and companies will aggressively bid, the role of the producer is to seek to lower one of their largest single line items.
Pre-purchase evaluation of visual effects is not possible, the VFX are effectively an intangible service at the time of awarding work. Visual effects companies bid based on the script or bidding materials, and while they may do a test, there is enormous variability, inseparability and intangible servicing of the directorial vision. The lack of a tangible product to judge and compare between is a problem for producers and VFX suppliers alike. This intangibility means producers are forced to judge on a range of quality indicators which are secondary to the final product and bundle the work into a unit that competes much more heavily on price than if the final product could be viewed, tested and evaluated before committing.
For VFX suppliers the intangible product issue increases risk, as they struggle to produce what is wanted, or even learn (divine somehow) what is meant by various notes often handed down second or third hand from the source (director or studio to production supervisor to company supervisor to artist). Not being able to meet expectations has a unique problem in the visual effects business. Given producers can not gauge the final result, they can be heavily influenced in awarding work based on how well they were serviced by this firm on a previous project. As the film industry is dominated by only seven studios, VFX suppliers are mindful of the need to meet the unknown expectations of the director and the studio, or risk losing the next project. With a very limited number of buyers and a very large amount of suppliers, the intangible nature of effects makes it hard for suppliers to dictate when the client has moved past a reasonable level of revisions, and additional fees should be charged as overages. Trying to charge overages too soon may seem unreasonable to a buyer who has yet to see anything matching their expectations, and who will be asking for bids on other work, even as this project is still in production (post-production). For example, R&H’s three largest customers, Warner Bros., 20th Century Fox, and Universal Studios, represent approximately 97% of the company’s gross revenues over the last three years according to court documents that were released.
Strictly speaking, VFX work is not normally actually a full fixed bid, in fact “the Studios have frequently said they don’t Fixed Bid, and technically they’re right,” explains Grossmann. “They get a bid, and then they approve overages after a process of natural resistance. So to their defense, they almost always change the final cost of the work to match the scope as it changes, but what they’re really trying to do is contain it to a ‘budget figure’ through horse-trading”.
The term is actually in part a vendor issue, VFX providers are often not being fully transparent about where the costs are coming from (either through inability to quantify, or from an unwillingness to share that information – see below.)
If one is to look at the bid from the studio’s perspective, then the studio should be saying “things have changed, tell me how that impacts us.” And the vendor embarks “on a slow, complex process of trying to figure out “what’s different about the assumptions and projected work involved with the change, and what money has been wasted so far on things that no longer apply.” Both of those are hard enough, but are made harder by the reality that they probably don’t even know how many man-hours actually went into wasted work, and also probably don’t have a clear vision for how much work will be required in the new direction,” explains Grossman. Usually the process of “figuring this out with even a 75% confidence takes longer than the vendor producer has to return the ‘change order’ and so a collective ‘ballpark guess’ materializes, and is probably negotiated and agreed to,” adds Grossmann.
VFX work is, of course, funded by the studios who commission the work. While it is the artists who do the work, it is the VFX facility that both manages the pipeline, often times the greatest aspect of the company’s IP, and they also manage the financial bidding, quoting, producing and billing of the work. For many years the preferred model has been a fixed or at least rigid quote, meaning that the work is not hourly or materials plus costs, but bid out on the basis of either a script, storyboard, treatment or previsualization of the film or parts of it. Unlike almost every other part of the film making process, the VFX company or facility is only paid on delivered work, and like only the director and the producer – they are not paid if the work shifts in schedule, is delayed or work cancelled after an agreement has been reached.
The analogy of a house construction is often cited. The VFX studio agrees to build the house using the land provided, for a fixed bid, but suffers adversely if the building is delayed, ground preparations reveal more work is needed, or if the rough design of the house changes. This last point is very complex to appreciate. How can a bid not be open to creative changes? The answer lies in that fact that the bid is asked to assume the ‘blue print’ is roughly, not exactly, what is required and that given the process is both fluid and creative – ‘some changes are inevitable’.
The problems facing visual effects companies are also digitally embedded. While less frequently discussed, the move to digital visual effects, and the age of rapid high speed digital file transfers has lead to a reduction in barriers to entry. It is possible as happened on the Oscar winning Hugo, where Pixomondo did a large amount of visual effects. As fxguide documented at the time, the workflow the company used allowed a shot to be worked on continuously around the world. One extreme example cited in the fxguide story was a shot turned over in New York from editorial to Pixomondo. An EDL would be sent to their LA office, where material was pulled and put onto the servers. The Beijing office would then start matchmoving the shot in their morning, and Shanghai would do modeling adjustments needed on the base geometry based on the onset LIDAR scans. By the time they were complete, the European offices would be coming on-line and the Berlin, Frankfurt or Stuttgart offices would contribute based on the skills needed (Berlin – effects animation, Frankfurt – character animation or Stuttgart for compositing). Later that day the London office would pick up the work and put it together into a shot or hand it onto Toronto. When it was done, the Los Angeles office would be ready for reviews and possible inclusion in the ‘days’ work to be shown to the client.
The same technology that allows one company to move a shot around the world, allows multiple companies to outsource or distribute work based on skills and going rates.
Digital visual effects, compared to traditional special effects, are also seen as being nearly infinitely flexible and thus able to be revised over and over again. If there are only two miniature models built for a practical explosion then to do a third take is clearly more cost and time. If the second revision of a digital explosion simulation is not right, it is almost always assumed it will be re-rendered or re-simulated until the correct creative output is agreed upon. The hard costs of equipment electricity, staff costs, air-conditioning on vast render farms, rent, and even depreciation are not seen in the same way as a crumbled miniature model on the sound stage floor.
Grossmann adds: “In physical explosions, it’s rare for a director to direct the nuanced details of an explosion and the way it moves because ‘It’s real.’ In digital production, because you CAN, and because ‘what’s real’ becomes a subjective debate, a director can transform ‘What he wants’ into ‘what’s real’ for the purposes of keeping the work going until time runs out.”
This situation is one of the VFX industry’s major time sinks and a real budgetary problem. Almost every VFX facility ends up with some of this on almost every show, of course, some directors are more notorious than others.
“The ‘what’s looks real’ vs. ‘what I want’ is huge,” says Grossmann. “Heck, a lot of the time we do things practically just because of that mental commitment that comes from the director when he shoots something practical. ‘What I want’ becomes ‘That’s what I got and I’m prepared to live with it’ as a subconscious acceptance in the director’s head.”
Artists work extremely long hours without job certainty, often without benefits or even permanent employment. A visual effects artist working in feature films in particular is likely to be an ‘independent contractor’, working as a global digital gypsy moving from project to project far away from the glamor of what one might call ‘Hollywood’. While there has been some movement on solving the illegality of such staff ‘re-classifications’ in some countries – it is far from solved globally.
There are many reasons for the problems and there is unlikely to be any silver bullet that will immediately or fully solve the problem, but it is clear a new solution should be explored.
While taking nothing away from the corporate structural problems, there are business problems that have been also discussed. Some of these include:
– complexity of bidding and shifting shots
– fixed bids vs issues of overages
– creative goals that shift and are ill-defined at the outset of filming
– shorter deadlines without scope for R&D
– the high cost of versions or revisions in zeroing in on a final shot (look)
Possible Solutions (or are they?):
Various solutions have been suggested for solving this, some solutions rely on increased subsidies such as the $330 million recently signed off by California Governor Jerry Brown, to match subsidies in Vancouver BC or Montreal in Quebec. Still other solutions seek to abolish or minimize the impact of these subsides, with import tariffs much in the way other industries have sought to adjust government endorsed trade imbalances – see below.
Equity in the VFX films: ‘gaining points on the backend’
Another often mooted suggestion is VFX companies seeking equity in the films they help make, or ‘blue sky back end points.’ The theory being that the firm would then at least share in the great rewards that blockbusters deliver. Unfortunately the track record here has been not good:
• Sin City 2 : Prime Focus (~$19M investment) – film did make make top 5 on its opening weekend
• Ender’s Game : Digital Domain (~$18M invested) – film only did $61 million US gross domestically
• Pan’s Labyrinth : CafeFX
• Yogi Bear, Hop, Seventh Son and The Golden Compass : Rhythm & Hues – none of these were huge B.O. hits
• Rush : Double Negative – film only managed $27 million gross domestic US
ILM is believed to have investment in Rango – significantly, that is a fully animated film with a domestic gross of $123M US. As will be discussed below, animation might be considered very different from live action VFX.
Thus far, an equity solution has not proven to work for a VFX studio, and it is worth noting that these equity considerations are not gifts, they are given in place of other payments, thus even if they were returning the investment made, they would represent a major cashflow delay from completion to payment. In effect, the VFX company is financing or bankrolling the film, in the same way a bank would on a loan. At its core this requires deep pools of capital, and the ability to spread that risk over a portfolio of committed projects. That’s the heart of all Hollywood film financing, and it is something VFX houses have rarely had, if ever.
While trade limitations would ‘level’ the playing field, “tax credits and subsidies have made the issues of the VFX industry really hard to ‘see.’ They really prevent global “apples to apples” comparisons, explains Grossman. “Companies in tax incentives areas might be significantly less efficient at doing a certain type of work, but be significantly more profitable because of the incentives. And what’s more challenging is the “soft-money” component: Even though I might get the work done cheaper in China on the bottom line, the decision might be made to send the work to a tax Incentive area because I can sell the tax credits in advance for 80% face value and use that capital to pay for production costs.”
Trade tariffs, if imposed, would certainly result in counter-moves by the studios to mitigate their exposure. Simply having tariffs wouldn’t make the “end of the day” picture any more clear for the people trying to make viable vendor financial models. This issue speaks most strongly to the movement of visual effects work around the globe, and the unsustainable nature of an industry in any one country if built on subsidies. Once the subsidies are lost or bettered then it is reasonable to expect the work to move to follow the better economic model. There is little research to suggest that a temporary substantial subsidy will enable a growth industry after the loss of the incentive.
Why would the loss of an economic incentive or advantage not logically be a disincentive to produce work in a region. If an incentive is assumed to attract work, logically the reverse must also be true, given the technology involved and lack of barriers to global entry/global work location.
However no matter what happens with tariffs there will always be currency, cultural and political reasons why work will not be able to be produced internationally and there the issue does not address many of the other structural problems. It may be a great issue to address but alone it will not make VFX a more profitable business, it can only remove market aberration.
Much like trade tariffs above, the case can be made that organized labor would be able to better service and protect the needs of a group or workforce, through collective bargaining and by removing the marginalization of any one individual who has tangible labor issues. But just above, this will not solve the structural problems of the industry, it may address fairer working conditions or more standardization but it will be geographically confined as most if not all unions who work for specific labor conditions do so inside the national labor laws of their home country. Unions may align internationally but it is much rarer to have labor disputes in one country spill to organized labor negations in a second country, it is possible but for sensible reasons of charter and membership, unions tend to be nationally bound. This is not to deny that there are many issues in the VFX industry and the USA in particular that might be immensely helped by unionization, health care benefits for example is one important such issue.
A third possible solution that is not mutually exclusive is the formation of a trade organization which will advocate a more balanced relationship between suppliers and the studios. A trade organization can take many forms, some resemble unions at a company level, others lose alliances with little in the way of explicit power to negotiate for concrete change. Such a trade organization may also face the issue of national boundaries such as when the VES, a nominally international body came out in favor of Californian subsidies against the interests, it could be argued, of its non-Californian members not only in places such as London, Canada, New Zealand etc but also their inter-state members in places such as New York etc.
Given the large number of potential suppliers and the ‘buyers’ market of limited major studios, it is unlikely that many of the structure issues involved in bidding, revisions, margins and sustainability could be solved entirely by any trade organization.
Again the problem is complex and any of the three solutions listed above – unions, trade organizations, tariffs and perhaps even equity – could all contribute to improved health in the effects business, but there is a very different approach: a radical change in the business model.
The New Industrial manufacturing Model:
VFX Visual Development
How is it that some animation companies have prospered while their VFX cousins have done so poorly? Clearly there are patchy issues in animation – Dreamworks Animation has had a run of films that have been lackluster, while at the same time Disney Animation along with Pixar have prospered. But on balance the CG animation industry is thought to be in better shape than VFX overall.
The concept may seem radical at first, but if you bear with it, the new system will seem almost obvious.
The fact is that filmmaking has been moving to more of a manufacturing production model anyway. Effects are now common place and yet as visual effects artists, there is a seemingly automatic rejection of considering our work as being something that can be done inside such a model and still be creative. But what if that is not true, effects are extremely complex but so much more of the gamut of effects leans to evolution than complete invention. Once could argue, the hardest thing left in VFX today is making a profit. If we assume then a manufacturing model would work, what would need to change?
The easiest answer is think of a creative large scale manufacturer of high tech good. Let’s pick Apple Inc.
Apple is an intensive manufacturing company, yet still a creative company. As we saw last week, the company innovates, it is stylish, it is profitable and it uses a manufacturing model for doing business. Apple ‘makes’ things, but it uses other companies to do it, but it owns design, prototyping and quality control. It does this by solving the creative problems and then getting companies to make X million of say this new prototyped product. It does not contact a bunch of outside companies and ask them to bid on some new type of phone or watch that no one quite knows how it will look yet! It does not ask for ballparks to make an iPhone device that is amazing and incredible and will sell millions but no one can even accurate draw yet. It makes a phone, or it makes a computer, and it solves design details. It gets bids and contracts to meet those specifications and Apple works closely with outside component suppliers: (graphics chips, CPUs, SSDs, etc.) to inform their internal R&D.
You can only hold suppliers to detailed specifications if you have solved the design details, sure a problem may come up, or a part may change, but then this would be solved by Apple’s engineers and designs with the supplier. The supplier would not be expected to just redesign the next iMac until Johnny Ives liked it, or until Tim Cook had run out of time before the launch date and thus had to ship it as it was. Apple investigates problems well before, way before – anyone sees a final production Apple product. And critically it knows exactly what to expect, the costing and process are not opaque to Apple. Apple knows how the darn thing works, it has manufacturing PhDs on staff, it does not just hope the next product will connect with audiences, it operates to be successful, and works with its suppliers.
Interestingly, Apple, who through its suppliers employs hundreds of thousands of people, managed to have 95% of their supplier’s employees work less than Apple’s allowed maximum 60-hour workweek (the average was 50 hours). Would that be true of 95% of VFX workers on major Hollywood films? Yes, many of those Apple suppliers are in China (349 suppliers), after all, Apple manufactures consumer electronics, but 139 are in Japan, 60 in the USA, and 42 suppliers in Europe. Not that there is a direct 1:1 comparison to be drawn between consumer electronics and visual effects, but some of the parallels are interesting.
How different would VFX be if, when bidding work, a VFX house was given an example of the final look to match and work to, recommendations on how this should be done, the details thought out of how it will work in practice and no ‘but it could all be different depending on the shoot’. What if one could bid the actual number of shots, and the actual complexity of them, using whatever pipelines you have in-house but with a very clear example of exactly what is wanted?
Unfortunately, it is not that simple. For a studio to do that, for a movie to be produced in this way, the VFX line item on the budget would not be opaque to the producers. For this to work, the VFX house – like the Apple supplier in real world manufacturing – would need to discuss actual costs per unit, and justify detailed costings. Not that this is unprecedented, the camera department do a budget with the rental company that the producer can review down to the tripod. The wardrobe costs are not packaged as a ‘deal’, the lighting budget often includes lighting gear, staff and even the burn times on the lamps. But the upside is, if you ask for an additional Softsun 100K light, the producers pay the $1500 fee and they don’t expect to get three additional lights just in case the director changes his or her mind. The lighting department also itemize their crew costs and their overtime costs. This is not just a union issue, although in some markets the union does actively help enforce such issues, it is more a product of setting up the project in pre-production and committing to what is wanted, while allowing the producer to decide that in fact they will have no Softsun lights and if that decision is wrong, the gaffer will not produce four anyway based on ‘fat’ or wiggle room in their department’s budget.
This also means that each manufacturer for Apple, or each gaffer truck, gets to compete on a level playing field. The desired result is known and the budget is itemized, broken down and transparent.
The new model introduces a more advanced visual development team around the studio VFX supervisor and or director. “Importantly, it’s a team that they’re paying for, so there’s no conflict of interest. The producers can weigh in about the expected total budget that the film can support, and the creative can explore in any direction they want to, within the limited set by their producer, not by a VFX vendor,” explains Grossman. This new entity would develop projects on a development budget and determine if the look is desirable and cost effective for the plot and target audience for the film, before principal photography.
Once visual development is done and agreed on – the actual work can be bid. Say three VFX houses might be invited to bid and they would be working on “here is what you should build, here is what you should do digitally, and here is how much we think it should cost (they know as they just did the shots to hero finals before bidding)…plus here are the assets, here are some production tools or scripts we developed during the visual development process, this is what we want, this is not a vague idea of what we want, this is what we want and this is how long it took with our people, tell us how long it would take at your facility, in this timeframe with this context? The bids back would be itemized and transparent, showing labor costs, hours and all fees explicitly. And most importantly, in conjunction with reels & resumes of previous work, this can be used to clearly identify where one vendor is more efficient than another at a certain type of work. No VFX vendors are equal,” explains Grossmann.
Why would this not work for VFX?
There could be many reasons, but let’s examine a few from the points above:
1. We need R&D to work out how to do these things and wiggle room to solve the creative problems
In this new manufacturing VFX process, this is solved either by the new form of pre-production/previs/design prototyping, or it is separately bid before the main bid is agreed to and signed off on. There is nothing wrong with a design company building rapid prototypes and proof of concepts that a third company will execute. Specialization such as this is common in manufacturing, but more importantly with previs we are already half way there, the sequencing is just not right, previs happens does happen but it is not aimed at solving or even costing final sequences. Previs currently serves a different purpose. “It has to answer two different production questions, ‘What do we need to tell the story?’ and ‘What do we have to build for for our shoot?’
“I would say that it also sometimes involves a third question,” adds Grossmann, ‘What does the VFX company have to build?’ but I feel like frequently that’s where previs falls apart and what needs to get built in post is not what was previs’d. Sometimes it’s exactly the same, a lot of times it’s not.” If your previs has no objective of making a blueprint for large scale production, it is hardly surprising it can only be marginally helpful in this regard.
Normally R&D, complex asset look development and creative changes have to be partially hidden in pre-shot costs. Yes, it is true a studio bid will contain ‘asset builds as a line item, “But in practice, this money isn’t really going to cover the R&D that’s actually going to happen,” says Grossmann. “It’s going to spill over into the shot costs, so that’s where a decent amount of it gets ‘tucked away’. Because if your R&D budget gets trimmed, you’re going to try to make up for it by trimming from your shot costs. That is also partially the reason why shots are packaged. To unpackage them a VFX company needs to:
• not fatten the per shot costs to cover inevitable changes
• not hide unsexy costs like archiving and asset development
• expose labour costs, and
• not have fixed bids.
2. Time. No studio will spend the time to do this
Actually while studios are relatively cost adverse,- some of this is already done anyway. Surely one could argue that a highly developed Comic-con teaser reel is very close to this model, with the exception that the bids are not tightly linked to the teaser and the VFX quote is not broken down or open. With hundreds of millions on the line, studios have already indicated that they want to see what the film will be like earlier not later.
In this model the VFX visual development phase would resemble script writing. Projects would be look designed before being green lit. Once the film is green lit and funded then time is money and it is a race to get the film made and cash outlays recouped as fast as possible. But before the film is green lit, this would help decide what should be green lit, how much it will cost, and how long it should take. It would be development money which is already being spent, and often on projects that are not then green lit and filmed. Let’s not forget films can kick around the studios for years before being green lit. Time is only short once the film is fully funded.
3. Big effects shops will never go for it
That could be true. This model may not work for the rarefied space occupied by Weta or ILM, – after all at this level the VFX houses are hardly guns for hire, they are more of ‘digital production companies’ (ILM + Star Wars / Weta + The Hobbit) but for every say MPC who perhaps might not be interested, there are a dozen of other smaller companies who would welcome no doubt a low risk bid and work schedule. This model would also allow for specialization. If you break down the effects before the film is green lit, you can farm out specialist shots to specialist companies. At the moment there is risk in this, if the shots change a lot then the very specialty that was quoted for, may not make the filmed sequence, but in a more tightly bid system it would, and if the shots were cancelled, there would be a cancellation policy. The work is no longer a rough guess, it would be a refined bid.
4. It does not fix subsidies
It does not. If a government wants to pay for an industry, it can normally distort the free market with heavy subsidies. Just like subsidies to film principal photography in Canada and double Vancouver for New York. If the subsidies are there, companies will work to take advantage of them.
But it does make it more transparent where the subsidy is having an impact and where it’s not. And it might provide some relief to people that aren’t in tax incentive areas because you don’t have to be doing “the whole show” in order to be doing an important part. For example, says Grossmann, “the studios I’ve spoken to have indicated that they’re willing to forgo the rebate in order to work with a smaller team of people close to the director, or specialized in a specific style or technique, or in development, because the savings wouldn’t amount to that much. It’s the volume shot production where they really need that rebate to kick in.”
5. Producers don’t understand VFX
It may be true that most producers might pull up on the details of spectral rendering or sub surface scattering or complex motion capture face rig editing. But they can understand labor, electricity, capital costs and much more, and “’VFX People” or ‘Digital Production Artists’ are surprisingly similar to their ‘Production’ counterparts,” jokes Grossman. And do senior film producers need to understand SSS algorithms any more than they understand the contents of hairspray or the reason this light needs 3 phase and this other one does not? Is it not enough that they see what these lighting packages do…and this is the cost of that? Producers, -real producers-, are wizards at spreadsheets and budgets, as much as we pretend our craft is ultra high tech beyond mere mortals, the basics are the same as any of the other crafts. Few producers understand why the LED spectral profile of the cheap lights are not going to work, they just hire good professionals who say that the cheap LEDs will not produce the right look, and who are willing to shoot tests to verify that if asked. Could we not do the same?
If a producer understood that re-doing a shot was $50,000 not because we just say so but because they can see the Deadline equipment summary on screen that shows the X hours of animators, computer rental, disk space, electricity = $52,450 (and here are the rates that went into that equation), don’t we think they would feel better about the $50,000? If you can not break down a number it seems arbitrary. No producer is just told that the shoot is $250,000 a day, they get to see where the money is going and use their professional judgement if that is going to make the best use of their budget in bringing the film to the screen. Maybe that $50,000 will make all the difference or maybe half as much and an extra stunt scene would be more effective? But at the moment the figure is just $50,000 and the only way to reduce it is to ask, beg or demand a discount.
6. It will cost more to prepare than it does now
Actually at the moment with the schedule of a film – when fully green lit – as crazy as it is, most of the start of a film is taken up with putting out fires and tying together information. If you have been involved in film production meetings as a VFX person before filming, it is answering questions, trying to collect and find out what is happening and when, and finally trying to understand what is wanted. There is little time to suggest ways to save money, or suggest ways to work more effectively, partly as it can take too long to go away and answer how long and difficult a VFX solution would be. The challenge for VFX is the pre-production phase is that there are normally just two VFX representatives in the room, a supervisor and a producer, and it is hard for them to come up with numbers quickly for a complex challenge, that usually gets done’later’.
“When I’m at my most effective for the overall budget,” says Grossmann, “is when I can look at a situation and say ‘how much is that gonna cost you to build/shoot practically?’ when I can in my head know how much time it would cost to do it digitally. The easy example is when you’re on set and they ask you if you need to hang a green screen in this new area the camera is suddenly seeing, or if you can roto it. I know how much our set day costs per minute, and I know how much VFX artists cost per day, and when the gaffer and grip come tell me it’s going to take 20 minutes to hang a green, I can translate that into man days of roto and say ‘we need to do this’ or ‘forget-about-it, we’ll roto. Let’s shoot.'”
It is crazy and the dominant thought can be ‘we’ll fix it later’…some VFX producers have commented to fxguide that there is no time in a typical eight to sixteen weeks of pre-production to do much more than hire coordinators, sort out computers, offices, get concept art, and responding to every other department.
“It becomes a receiving department rather than a pro-active department that goes out and works out how to save the producer money,” explains Grossmann. “There is no time to say “here are the assets you’ll need, here is how they will look, here is how to assemble them.” The VFX team in preproduction are left saying “well yeah sure we could probably do that and here is the least destructive way to do that”.
And then once the film has been shot, after the imagery is committed to, people will work out exactly how to make a shot work, even if that is perhaps not the best shot or the best approach. This can translate into re-shoots, and it most often means unexpected and most often unfunded, overages or re-dos for VFX vendors.
This new plan involving visual VFX development would allow producers to see that maybe an idea on paper was not working on screen and fix it in the script, not in post.
As VFX can solve so much and do so much should it not be given a chance to work its magic before principal photography? The idea of ‘visual development’ before a film is green lit could save not only money but perhaps help a film be made better. Producers need to see the raw material costs, labor costs with transparency. VFX houses need exact bids and work like a manufacturing supplier.
For non VFX producers the question must be “where does the money come from in this new structure?” Grossmann says it happens every time. “They love the concept, they agree with it, it feels very traditional, but it’s new. So they ask how money flows in this system”.
Grossmann says the team has tested the idea of making VFX an extension of other departments in Prep. “So in one case, we were billing through the Locations department budget for concepts to further locations they were considering, and for previz that we did to show both on-screen visuals and logistics planning. We’ve also billed through the Art Department for development work that was just beyond the scope of their capabilities, and because of our capability and proximity, evolved into using the VFX team to do design work and planning that wouldn’t even involve VFX in the final implementation. It’s very cool to see the barriers between the departments come down like that. I can already see where these departments will have their own “VFX capable people” in the future”.
It’s easy to say that there should be a separate visual development budget much in the same way there’s a story development budget. But the question that’s come up that’s harder to answer is “I get that logically it makes sense to spend money up front in order to save money on the backend, but making that happen will not be easy.
7. The director will hate it
That is possible, but it was the director who created the look and style, so for some once they are on set it would be very helpful to know what they were doing. It would be helpful to be filming to their vision a reality, and not filming with the only the hope that this will turn out well. In the past some directors liked detailed storyboards, some did not, but for those that did, it can mean they can focus on the acting and the performances and not on the complexity of guessing how things might work technically.
“So far, with directors, the summary I hear is ‘You mean I get an Art Department with full VFX capabilities early enough on the project that I can spend time with them experimenting on the film, even on things that might not become VFX in the end?’ What would I not like about that?,” says Grossmann, from first hand experience pitching the concept to a range of senior experienced directors.
Directors can now rely on their own judgement – made at a time when there was time, and not in the middle of a midnight shoot on location in the cold with a vast army of people wanting answers and the move on to the next setup.
8. Money = labour
It has been pointed out by more than one key industry expert that the elephant in the room for this discussion is that Labor costs are the most significant part of the VFX expense. People say 70% of the costs are directly spent on labor, fxguide knows of projects where Labour accounted for 80% or more. But ultimately this approach and all approaches are really about trying to use labor more efficiently in order to reduce costs while producing great films.
The reality is that any artist reading this can attest to major wastes of their time on shows. fxguide hears about it constantly. People worked for months on something that got cut, or spent months more wandering the desert in search of creative on a sequence. Or waited weeks for feedback, etc. And all that amounts to artists wasting their time (which they don’t like regardless if they’re getting paid for it or not) and time costs money. Big money.
Because it’s such a big part of the “financial problem”, studios have sought to leverage market share to drive labor rates down, or keep them low. Obviously then, the easy way to try to have an impact on your budgets is to negotiate down the rates on the major item that accounts for 75% of your total cost. What’s harder is making that labor more efficient by giving them faster computers and resources and pipeline efficiency. And even harder still, by giving them clear direction and fast feedback. That’s a key part of this visual development approach.
Note that Time & Materials isn’t always good, and Flat Bids or “per unit costs” aren’t always bad.
There is no one solution to everything, and that is true here too.
Time and Materials can be tricky unless the labor and resources are tightly managed, “so you know exactly what people are doing, what they’re working on, and what they’re using to do it” explains Grossmann. In addition, knowing one’s overhead ‘exactly’ is critical to determine the edges or ‘fringe” costs in a time and materials model. “We’re notoriously bad at that in VFX. Plus, Fringe costs get “attacked” right away in a negotiation, so it’s hard to defend high fringe costs for a specific service you’re providing, if your overhead is being calculated inclusive of a much broader array of services”.
Flat Bidding isn’t so bad if one is competing head-to-head with other companies where the objective and comparison metrics are clearly defined. “This lets a company that’s “more efficient” reap the financial benefits of having lower internal costs than the “market price.” But it’s terrible when the scope of the work is vague or the objective tends to change drastically”.
Because all of these various scenarios are playing out within the structure of a “VFX Vendor”, it’s not easy to switch financial models depending on the scenario. But each movie can have both scenarios. “In essence they all kind of lump together to create a “risk portfolio” and once you do that, you can’t target changes in specific costs or shots, or sequences without jeopardizing the stability of the whole financial system. In essence what we propose is to take these different types of scenarios and assign them to different types of teams or vendors, under the appropriate financial models”.
Will this work?
It already is, or starting to – several films have already tried and liked it. As we write this MAGNOPUS is finishing an impressive special project short film for Director Martin Martin Scorsese, which they budgeted and then executed in a time and materials model, and took that approach with some of their vendors, where appropriate. They also did some flat weekly cost for a specified crew size at the vendor. All of it was an experiment to see fo the company to see what the challenges would appear with each approach. Ben Grossmann and the team at MAGNOPUS have been working with a set of Hollywood studios to implement this actual model. And they are starting on another time-and-materials project for a major VFX movie right now, which is very much a “development” project, and thus it lends itself well to billing actuals on labor plus fringe and equipment rentals.
The talent team at MAGNOPUS runs deep, Alex Henning and Rodrigo Teixeira are partners in the company and have been key to developing the ideas in this article and applying them to the company’s test projects. Craig Barron (formerly of Matte World Digital, & Oscar winner for Benjamin Button) is also a key creative; Magdalena Wolf, member of the Producer’s Guild, has brought a VFX producer’s perspective to the discussion, and there are many other senior artists on full time staff and working to explore the concept.
While MAGNOPUS is well placed to be a design effects company, they are highly trained and well resourced, they also know that for this to work, there has to be a set of companies working this way and an even bigger set of VFX vendors who are willing to be transparent and work with studios this way. Not every studio wants to convert to working this way, but nor has there been a lack of real interest in moving to a model with more transparency and real visual VFX development.
MAGNOPUS is also not just a VFX design company, they are servicing a range of clients outside the film industry will ‘VFX skills and VFX style services’, but they are also working hard to be the ‘VFX visual development evangelists’ to the industry. The team have worked both sides of the fence, supervisor and supplier, they know all the top studio production executives and many extremely influential directors. In other words, if anyone can pull it off these guys can, but from visiting their new loft production offices in downtown LA, they don’t even want that many projects, their game plan is wider and more complex than just this one role.
As for the VFX community many top specialist VFX houses have already heard the pitch and are considering it. The merits are clear but the downside of not changing is even more pronounced. If we are to grow, or even survive the current situation things need to change and perhaps this is the solution.
For the artists there are two sorts of work, the highly specialized teams that work with the director before the film is made, with all the advantages of not sharing their attention and needing to move fast and accurately. The second group is a focused group who are aiming to do exceptional work to an accurate brief, but without an open ended outcome.
VFX houses need to be willing to be transparent. Nothing is going to be given for free. If a studio commits to this model, they must be allowed to actually ‘produce’ and that means line by line cost explanations.
It will require some adjustment, programs such as ftrack or Shotgun will need to help and provide an efficient translation between productivity and cost (i.e. virtual time cards that are literally tracking labor to tasks) plus resource costing: so a producer can be answered immediately when they say “I know how much a 2×4 piece of lumber costs, but what does this render hour cost?” And critically the teams will need to very quickly access this information, often in real-time from on set or in pre-production.
For our money, here at fxguide, this is one of the most comprehensive and thought provoking ideas we have heard and it deserves wide scale industry discussion. It is possible to flood the debate with the sins of the past, and the injustices of the present, but we need to also be able to debate the merits of this possible future. No one, least of all fxguide, is walking away from the problems we have experienced, but we ask your input on what we can do to make it better for the next generation of artists. We owe ourselves and them a considered nuanced discussion, there is no silver bullet, but structural change may help.
Note: Mike Seymour will be chairing a panel at Siggraph Asia this Wednesday morning looking at the business issues in Asia. If you are in China at Siggraph please come by the Asian Expansion Panel Wednesday 9.30am
Experiences in Asian Expansion
COO / Executive Producer,
Founder and CEO,
Base Media Group LTD / Base FX
Head of Animations Studio,