The Fallacy Of Infinite Scalability

If you dare to unionize, the work will go away.

It’s the cynical argument I always hear.

Yet if you think about it, if it was that easy to send work away to a cheaper place, why does it continue to go to some very expensive places like New Zealand, Vancouver, and the UK?

Every now and then I often run into a Californian VFX artist packing up to do a year or 2 in the UK or NZ. I curiously ask how the pay is to which they jokingly respond:

“It’s basically name your own rate. They are desperate for workers.”

I often write that the real reason many of these areas are attaining so much work is because their bids are artificially cheaper through subsidies. National governments offer to directly rebate US studios for film work done in those locations for about 20-35%. That’s pretty substantial.

Even with those discounts, a significant amount of VFX work still gets done in California even without subsidies. Why is that?

Much of this has to do with scalability, the ability to handle a large amount of work at any given time.

Since VFX is always being designated as a commodity, why not compare it to other commodities like gasoline for example.

If the gas station across the street is offering a 20-35% rebate for you to buy gas there I’m sure you, I and every person around would make that gas station the first choice to fill up.

Pretty soon there would be a line around the block with long waits for gasoline at that particular station. The station across the street offering no discount will probably be empty… unless there are people with very little time on their hands.

That wrinkle is an important distinction with the VFX industry. Studios don’t have an infinite amount of time to wait for a discount. As in many cases like Green Lantern, the work eventually was done mostly at Imageworks in Los Angeles.

Now what if that gas station that offers the discount started running out of gas? Well drivers in line would probably try to gage the capacity and if things weren’t looking good, they would jump to buy the more expensive gas. Many facilites in the UK, Vancouver, and NZ turn studio work away because they are solidly booked up. It’s not like you can just walk in and have them make more vfx for you.

Let’s add one more wrinkle. When you finally get your gas with that discount, you are asked to pay in another currency: Australian Dollars. You quickly realize that your discount is evaporated by a higher currency.

I’m compelled to speak about these scalability issues because there have been a few articles about the growth issues in 2 heavily subsidized locations: The UK and Australia.

For the UK, they are realizing that they are heavily reliant on foreign labor that isn’t cheap:

A report, commissioned by the government and published earlier this year, delivered a worrying prognosis.

It warned that, while special effects was enjoying a rapid growth, the sector was also “having to source talent from overseas because of skills shortages at home”.

Well duh. Why not go to India or China? (I’m joking.)

For Austrailia, even though they are heavily subsidized, they suffer from a very high currency which makes it expensive for US studios to do work there. Even with deep discounts, workers are looking for stable jobs. In a project-to-project industry like vfx, you need to situate yourself where you can jump on to a few projects at different facilities:

“Projects like Legend of the Guardians and the Harry Potter series have allowed Australian companies to showcase world-class capability, but there’s not enough consistency of work to maintain staffing levels, so we’re all stuck in a boom-bust cycle,” he said. “This volatility makes it harder to maintain staffing, undertake artist development and remain competitive in the marketplace. The constant relocation and training involved in bringing new staff on a project-by-project basis creates an overhead which makes us less competitive against overseas providers. It’s normal for people to move, but those kinds of volumes are dangerous for our industry.”

His counterpart at Animal Logic, Zareh Nalbandian, agrees: “We need to offer people a sense of permanency, that they can make a long-term commitment, raise their family, buy a house and all those things that go with life.”

The irony in all of this is that subsidies, the very mechanism used to lure US studio vfx work, is the cause of much of that volatility. When you combine that with the fact that you are unable to quickly scale your workforce up for a large amount of work and have to pay sky-high rates to lure foreign talent compounds the issue.

So next time you hear someone say “It’s all going to…” stop em dead in their tracks and tell them “it’s all going to where the artists reside”.

Soldier On.

26 Responses to The Fallacy Of Infinite Scalability

  1. jai says:

    I’m curious – is there a sense that there is unemployment or underemployment of US artists – or just that they are subject to the same boom-bust and dislocations project-wise, which are described above for other countries.

    aren’t the best vfx facilities in the US booked up and hurting for good talent too – made harder by the H1B rules and the inability to do fulltime hires …

    are people saying the jobs are going away, or that the job conditions are majorly deteriorated due to global sourcing ?

    • VFX Soldier says:

      I think you could say its a little bit of both. The US vfx market has a very dynamic wrinkle to it: Many vfx artists work directly for the production companies like DreamWorks Animation and Disney. Blizzard and many games companies employ vfx artists with staff positions.

      This causes big problems for vfx facilities.

      When Imageworks let many of it’s artists go, they were sucked up by Dreamworks. In fact, The Animation Guild has record membership. This has caused some vfx facilities to have huge crewing needs when a big show comes along. So you could argue that there is underemployment.

      However, there are good spurts where everything is booked and some are left looking for a job. The longest I have been unemployed is 6 weeks and that is because I’ve turned bad offers away.

      I can tell you that we have huge needs but the money just isn’t there. Schedules are compressed and people are trying to do more with less. We are also having to TRY to outsource work to tax incentive havens. Eventually push comes to shove and the work comes back. Because of all of this, I feel the quality has gone down in my opinion.

      I think the general consensus is that if you have a job, you cling to it. If you don’t it depends how urgent you need a job. If you need one now, you can jump to another country. If you have experience and can wait (my rule of thumb is 6 weeks) you can get a job.

  2. You did notice that Imageworks Albuquerque office did most of Green lantern (65%) for the longest time and only during the final push Culver City ramped up artist because of the local talent pool.

    • blah blah says:

      is that why it looks like crap?

    • VFX Soldier says:

      Andreas, what’s the source of that 65% figure?
      Secondly, when you say the final push involved a ramp up in Culver City and not New Mexico implies what this post about: The fallacy of infinite scalability.

      I’m sure if the execs could do all the work in New Mexico (which was the original intention) they would do it. But they couldn’t because much of the talent wanted to stay in LA.

    • ed says:

      A shortage of skills is not the case as it applies to animation, there are so many animators that it comes down to who tows the political line and schmoozes the leads the most. I’ve seen animators of superior skill get ousted because they get vocal about politics, the reason given by the company is always ‘not enough work’ but it’s really because they rubbed some tossy, status monkey producer up the wrong way with a conspiracy theory or two on the email newsgroups. Some companies in the southern hemisphere tend to hire Americans and Europeans over locals because they’re still under the delusion that there is a local skill shortage and assume the talent is overseas, no matter how good the reel.

  3. Reality says:

    The idea of bypassing the facilities and dealing directly with the studios has been something of a recurring theme on your blog (and elsewhere), and seems to have its share of supporters. However, after reading this post, I feel compelled to point out at least one of the logistical nightmares that kind of production model would present. Part of the issue, I feel, is that most of its supporters are looking at it through the (relatively) rose-colored glasses of “people who don’t have to solve logistics problems.”

    I’m currently working as a pipeline TD (mainly Nuke-side), and I can assure you that the overhead mentioned throughout this post relating to project-based staffing swells would absolutely wither in comparison to the overhead of constructing an efficient, large-scale, functional VFX pipeline every time a show was started. You think being able to “quickly scale your workforce up for a large amount of work” is a challenge?!? Child’s play. Oh, and guess what? Now you have to populate that pipeline with artists who know the particular packages you’ve decided to implement.

    I’m sorry, but I don’t care how many talented coders you have on your project… you’re either going to be playing fast and loose with lots of catch-up, babysitting, and duct tape, or you’re going to need a sizable chunk of lead time and a nice, slow ramp-up to get things running smoothly. So yeah… good luck getting the studios to extend their post schedules.

    Compounding this is the fact that none of the developers and TDs would be particularly thrilled about having to partially or completely re-write, port, or otherwise re-purpose code and tools over and over and over again without any time to develop anything new and exciting. Can you say stagnation?

    Some of the greatest VFX tools in the wild today have their roots in in-house development at (say it with me) established facilities, where someone clearly felt they were in a stable enough situation to explore new ideas or techniques. Even if everyone else decided to go freelance and work nothing but studio contracts for the rest of their lives, I would put money on the devs and TDs staying together and forming facilities of their own. And if not, I guess there’s always India…

    I’m 100% pro-union, but personally, I like having a familiar group of people to work with, and a familiar code base to build on every day.

    • Dan says:

      Hear-hear!

      I’m glad someone has pointed this out in such an elegant way.

      The future of VFX is the facilities getting involved with the studios more closely, getting involved mopre thoroughly in pre-production and incorporating residuals into their contracts/bids. The facilities will have to take on a slightly bigger risk as they will only make money on residuals if a film makes a profit and all the complications that this will bring.

      Working for a studio will bring a whole new set of problems, problems that we have solved by setting up facilities. Seems like a step backwards to me.

    • me says:

      I hear this argument made constantly every time this subject is brought up, and I’ll make the same argument back that I always do:

      The cost of setting up a pipeline is peanuts compared to the costs of actual production. I was a pipeline TD at a few big facilities for many years before transitioning to the artist realm, and I can tell you that setting up a pipeline is not in any way trivial or unimportant but it’s also not impossible. The number of people working on setting up systems for artists to work smoothly pales in comparison to those artists actually working.

      Moreover, the idea of a ‘pipeline’ in general is something that gets tossed around quite a bit but is always discussed in very vague terms. Are we talking about a production tracking system? A render queue? A version control system? Those are all completely separate systems with their own challenges and solutions – some custom, some off-the-shelf.

      Bottom line – there’s some hard math that studios need to do before they decide to take on these challenges, but they should do that math if they haven’t already. We’re all very smart people in this business and work around whatever solutions will best fit, for today and in the future.

    • VFX Soldier says:

      As a pipeline TD or software engineer, how would you or the work that you do be limited by working directly for the studio?

      Secondly, in today’s practice, we have seen the studios successfully implement groundbreaking R&D. Look at Pixar, DreamWorks, and Disney.

      • Reality says:

        @me: I think you misunderstand my issue with the decentralized free-for-all work approach. I don’t pretend to compare the volume of work done by TDs and devs to that of the rest of the crew, but when you have VFX studio higher-ups lamenting the amount of overhead wasted on staffing up for every show, and then turn around and tell everyone they shouldn’t be working for a facility, it exposes a serious disconnect from reality somewhere in the thought process. Not only are you talking about staffing up on a much larger scale every time someone signs a check, but you’re also suggesting someone construct an impromptu facility ANYWAY to accommodate the show’s needs.

        I’m not talking about money. I’m talking about an excessive and altogether unnecessary time commitment in an age when schedules are constantly shrinking. And I’ll qualify this by adding that I’ve worked as a compositor, lighter, and FX artist, so I’m not just looking at one side of the coin.

        Yeah, sure, you can try setting up a 200-person facility using nothing but Maya’s core functionality and an off-the-shelf installation of Rush, and then pushing a 250+ shot (to be gentle) feature show through it. Good luck. Nuke is the only package that even comes close to being drop-in production-ready for a “wild-west” freelancer team straight off the shelf, and you’d still want to bolt on a few shiny bits before you (I) were satisfied. If you run and gun with that size of a team, you’ll be fighting your resources every step of the way.

        Since you’ve asked me (indirectly) to clarify the definition of a pipeline, here are a few of the things I would expect to have in place: reliable ingestion frameworks for motion capture data, on-set captures (i.e. LIDAR scans), and other production assets; asset tracking and version control; seamless asset, shader, and rig/anim cache interchange between layout scenes; automated convention enforcement; a flexible render queue with the ability to add arbitrary job classes/procedures; a reliable transcoding/mastering backend for outputting dailies, client reviews, editorial deliveries, etc. Oh, and a mini fridge.

        @Soldier: In response to your follow-up comment, the work that devs and TDs would have to do wouldn’t necessarily be limited… I meant to imply that it would become mind-numbingly repetitive and frustrating (more babysitting than innovation). Part of the reason I’ve shifted from comp to TD work over the last few years is the number of fresh challenges I get to take on every day. That’s the part of my job that has always gotten me up in the morning.

        Yes, some studios have implemented “groundbreaking R&D.” But those are all (here we go again) established facilities. How many revolutionary VFX tools have been developed by random groups of otherwise-employed developers?

  4. David Jones says:

    If I read your post correctly, you appear to be saying that American VFX workers need not be afraid of jobs disappearing because:

    1) Foreign studios will eventually become booked; work will overflow back to the States.

    2) Currency conversions will keep some of the foreign government incentives in check.

    To which I reply:

    1) I think this becomes less of an issue as more and more studios open shops in tax incentive areas (Pixar, DD, MPC all opening shops in BC, etc).

    2) Good luck pinning your hopes on your local job through fluctuations in the world economy.

    But the question I have regarding unions, that no one ever seems to address is this:

    What makes the VFX industry of the present any different than the cel animation industry of the 40s and 50s?

    Cel animators were unionized, and that didn’t prevent cheaper work in Asia from destroying the American animation industry. To me, the writing is on the wall. Some vfx work might remain in the States (just as Disney maintained some cel animators into the 90s), but ultimately, the industry will move to cheaper shores, union or no union.

  5. David Jones says:

    Actually, my argument is that REGARDLESS of unionization, the work will go away. When vfx companies overseas (eg: India and China) start to compete aesthetically with local vfx firms (and no one should doubt that’s going to start happening in the next 10 years), small to mid-size firms will be unable to compete.

    I predict that only a few large scale vfx firms with huge RnD budgets will be able to compete (eg: ILM) without government subsidies.

    I think the union discussion, while not unimportant, detracts from a larger issue regarding the survival of vfx in the States.

    • would that not be the case for everything? I eletronics are made in taiwan or japan or china. but i still see local businesses do a lot of work. how come? convenience, craftsmanship, I dont think all work will go away.

    • VFX Soldier says:

      Well wait a sec didn’t you write this in your original comment as an example that eventually work goes away?

      What makes the VFX industry of the present any different than the cel animation industry of the 40s and 50s?

      If that were true, the trend in The Animation Guild membership would be down. But if you look at the evidence cited, it’s grown: to a record high.

      I can understand the idea and the rhetoric behind the “work will all go to china and india” but it’s just not happening with vfx. It’s going to go where the talent resides and right now a good chunk are going to the UK and NZ because they are paying some really good rates for foreign experience vfx workers.

      • Even though i am in NZ currently I find your lack of believe in the CA Talent base sad. CA still has as many vfx artist as UK and NZ, maybe a little less buts by no means a small industry and those jobs are staying here. Sure some died with Asylum, the O, cafefx and the others but overall I dont feel our location weakened.

      • Mike Vargas says:

        One of the big issues facing competition with Asian studios is crack.

        http://www.fxdag.org/blog/2011/06/16/competing-against-studios-on-crack/

        When I’ve asked Chinese and Indian counterparts about their software licensing, I’ve been quickly assured that it’s all above board because they only use “open source” versions of Maya, Nuke, Fusion, Softimage, Photoshop, whatever downloaded from the internet for free.

        Unfortunately the Chinese and Indian definition of “open source” is pretty much the same as the American definition of “software piracy”. The use of “Crack” across Asia is epidemic.

  6. […] If you remember, I wrote a post about scale. […]

  7. […] Now the subsidies in Canada are meant to be used for Canadian workers but with so many vfx studios opening shop I’ve heard a producer say they can’t find workers. Remember my post on the fallacy of infinity scalability? […]

  8. […] the state of the union in vfx…  You can read about vfx as becoming a commodity here and here.  Hard work, unfair condition, pressure from clients in an industry with very few buyers, vendors […]

  9. […] #VFX and The Fallacy Of Infinite Scalability […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: