Back to: Introducing ... Death Knight Love Story, evolution of Machinima | Forward to: A Virtual Filmmaking Primer

Why Make A Movie Within A Computer Game?

So, one question that a number of people have asked about Death Knight Love Story (DKLS) can be paraphrased as "Why Machinima?"

In other words, "Why make a film using a computer game? Isn't that, you know, a bit crazy?"

Now, when it comes to the decision to set the film in the storyline of World of Warcraft, I might in hindsight agree. Previous films of mine have used game assets without touching the game's storyline, and whilst I was attracted to the idea of Rozencrantz and Guildenstern Are Dead-ifying the World of Warcraft, in hindsight we'd probably have been better writing an independent, non-WoW-associated plotline.

But as for using the game itself - no, that was a spectacularly good move. A move which enabled the film to be made at all, in fact.

What IS Machinima?

For those who are unaware: I'm best known as a pioneer of a subgenre of animation called "Machinima".

The definition of "Machinima" has teetered back and forth over the years, between "filmmaking in a 3D environment rendered in real-time" and "filmmaking using computer games". The debate has raged fiercely in the Machinima world over the years, and many Machinima purists, including me, have at one point or another gotten very irritated at the way our artform is so strongly linked to gaming.

However, I have come to realise Machinima was and is the result of the remarkable potential of video game worlds to enable a live-action filmmaking style within a virtual world.

The Astonishing Cost Of CGI Animation

Conventional CGI animated moviemaking is infamously slow and expensive. In fact, it's arguably the slowest, most expensive narrative medium in existence, challenged only by stop-motion animation. And one of the key reasons why that's the case is that for each movie, the entire world in which it's set must be created from scratch.

That sounds pretty intimidating as a task. It's actually far more intimidating than it may appear.

Take the characters of the film, for example. Obviously, each protagonist, antagonist and other speaking role needs to be sculpted from scratch. That's a task with multiple stages (concept art, rough sculpt, texture painting, fine sculpt, facial expressions, and matching the character's skeleton to the body in such a way that it doesn't deform unnaturally), and it's not cheap: $3,000 - $5,000 at an indie level to produce characters like our two protagonists in Death Knight Love Story.

So, it's possible to look at a pared-down script and say that, assuming two protagonists and two antagonists, we'll need to spend minimum $12,000 on the task.

Oh, but wait - we also have incidental characters. Any story not set in a deliberately extremely confined location will need quite a lot of these characters: read through any of Charlie's books and notice how many incidental characters have actual or implied lines within a few scenes.

In the case of Death Knight Love Story, the characters mount up quickly. The two traders at the start of the film, the Captain of the Guard, the head of the cannoneers, the other Death Knights in Naxxramas, the adventurers (4 of them) - and that's all within the first five minutes. There are at least another 15 of these incidental characters through the 20 minutes of the film. They need somewhat less attention to detail than the main characters, but if we were creating Death Knight Love Story in a conventional manner, that's still another $30,000 or so needed in the budget.

Well, at least we're done now. Except we're not - because this is a visual medium, and that means we need to consider the background characters needed to make our world feel like a world, not an empty set. And those guys and girls mount up fast: 20+ within the first scene of Death Knight Love Story, far more subsequently. Some of them can be copied and pasted with minor details changed, but there's still a lot of work to do: if you were to assume another $60,000 minimum devoted toward background characters on a film the scale of DKLS, you'd probably be well on the low side.

Now, *looks at smoking chequebook meaningfully* shall we talk about sets? Or props? Skies, backdrops, plants and animals?

The summary of all this: creating worlds ab initio is absolutely standard for 3D animation, and is also really, really goddamn expensive. And so, even if you've got access to motion capture and all the clever tech tweaks that we used on Death Knight Love Story, the cost of hammering together the world in which you will eventually create your film makes animation completely inaccessible for anyone with less than a mainstream TV budget.

Unless - as we did - you have access to a virtual world, such as those created for modern-day computer games. In which case, all those costs go away.

Making Films In A Virtual World

If you have access to a virtual world like Azeroth (the world of World of Warcraft), you can suddenly take a live-action filmmaker's approach. Rather than creating sets, you can simple location-scout. You can choose a landscape, a castle, a house suitable for your project from hundreds of those locations all over these virtual worlds. The process becomes very similar to that of a conventional cameras-and-real-actors film: shortlisting possible locations ("How about this house at the back of Stormwind?" "No, too dingy - how about this one in Goldshire?"), picking one, then moving the cast and crew there.

The only difference is that in real life most directors don't have access to a "/teleport" command.

Rather than creating characters, you can use the existing population of the world, too - most of whom are very happy to take instruction, thanks to being run by an AI script rather than an actual person.

(More or less, anyway - some Machinima makers, particularly those using The Sims, spend good long while wrestling with the problem of persuading their actors to actually cooperate.)

Furthermore, rather than sculpting your lead characters from scratch, you can use the character creation tools provided in the game. Virtually all RPG- or MMO-style games offer a massive range of "casting" choices in the name of making your character unique.

Creating a tool for desiging well-sculpted, appealing protagonists who fit into the same consistent art style is surprisingly complex. However, the number of people out there who want to play an AAA computer game is massive - easily enough to justify the investment in a tool capable of churning out almost infinite numbers of beautifully sculpted, world-consistent protagonists.

And so, suddenly, for an independent storyteller, the projected costs of any independent animated project plummet if they're willing to use a "Machinima" approach.

Of course, their story has to fit within the locations and characters offered by a game's world; but there are so many games out there, and each of them have such vast, sprawling worlds that it's comparatively easy to find a good prospect.

It would be possible to tell almost any low to mid-fantasy story using Skyrim or World of Warcraft, with some minor modifications. The genre of science fiction, at least in its mainstream film-friendly incarnation, is similarly well-covered by EVE Online, Half-Life 2, Dead Space and dozens of other games. (It would take some thinking, but I could probably make a fair stab at adapting any of Charlie's SF works using one, another, or a combination of game engines.) And of course, modern-day stories are thoroughly covered by everything from The Sims to Call of Duty.

Now, the chances are that you're wondering about the legal aspects of this usage - and that's an entirely different kettle of fish, and a rapidly-evolving one at that. ( A situation I can assure you isn't particularly comfortable if you're one of the fish. )

But in principle, this is why Machinima is so attractive: because it allows animators, rather than having to create their worlds, simply to become live-action filmmakers and take their cameras into worlds that don't exist.



One thing I think is interesting about machinima, is that in WoW, people use straight up game models and animations most of the time. (Exceptions like Craft of War: Blind exist, but yeah.) Meanwhile, in Minecraft, there's actually improved versions of the 'Steve' type character with eyelids and more joints and all sorts of changes. I actually think I see more machinima with those than I see videos like Steve vs Steve that use the in game models. Is it a difference in audience between the games? Is it a difference in the content I personally have been exposed to? I don't know.


That's a very good question!

It's largely down to the complexity of the models, I think.

It's very easy to take the Minecraft models and add additional complexity - I've done so myself, and the model modification there was a matter of minutes rather than hours or days.

By contrast, the WoW models are much harder to alter than you might think. Their skeletons are fearsomely complex, with 80+ bones in them, and a lot of the rigging decisions aren't exactly what you'd expect. It took me a good day or so to figure out how the base rigging works on a WoW character, and I'm pretty good at this stuff.

And whilst their faces are comparatively low-polygon, they're complex enough that altering them requires Proper 3D Modeling Skillz, not just 5 minutes with the "Box" tool.

Plus, the tools that exist to extract WoW characters are finicky and user-supported, and the tools that exist to extract the sets and maps are even more obscure. I believe the tool we used to get the WoW landscapes out of the game isn't actually available online at all any more.


Thanks for the cost breakdown, it's really interesting to see. I had generally thought that CG was cheaper than traditional animation and that's part of why it took over. How would you say that the costs of CG compare with 2D animation? I'm sure there's a massive difference between Hanna-Barbera and Disney, but a general idea would be helpful.

I'm a video game fan, and once of the reasons people bring up for why we don't see pretty 2D games anymore is that it takes a massive amount of time and effort to create good sprites over models. Sprites are easier on the low-budget end, but once you hit about the PS2 era it's just too much effort for anything but specialized studios like Vanillaware. I'd guess the same is true for movies, but I don't know. (I don't even know if what I've heard is accurate)

Of course, a big reason for 3D over 2D is that people just seem to like 3D animation better. I personally feel otherwise, but it's not really cost-effective to make something just for me...


2D animation is actually cheaper on very short productions. It's quicker to draw a frame than to create a 3D character, by a long shot.

If you use limited animation techniques, and particularly if you also incorporate 2D keyframing techniques used in software like Flash, you can actually get a long way before the cost of 2D starts to overtake the cost of 3D. I believe (although Nojay can probably correct me on this one) that's one of the reasons that a surprisingly small amount of Anime is 3D animated, even now.

However, once you've created the characters, 3D starts catching up fast. It's a lot more flexible, and certainly animating 20 seconds of movement on an existing 3D character is far faster than drawing the frames for those 20 seconds, even if you're hand-animating.

Moving away from discussing purely hand-animated scenes: as soon as you add in motion capture technology, the production time drops like a stone. Given pre-created models and a preexisting scene, I can shoot scenes faster than an equivalent live-action set, never mind an animated one. The Minecraft short above took approximately 10 minutes to film and 6 hours to take from setup to release.

However, mocap has a completely different visual aesthetic to hand-created animation - that's why Pixar don't use it. It lacks squash-and-stretch, and many animators have argued that it shouldn't even be referred to as animation at all.

Personally, I have a certain amount of sympathy for that point of view. I've always thought of myself more as a live-action filmmaker who just happens to do most of his production work in places that don't exist.


many animators have argued that it shouldn't even be referred to as animation at all

They must really hate Bakshi-style Rotoscope then.


Oh, yes. People got seriously upset about rotoscoping back in the day.

And these days, Mocap is occasionally referred to as "The Devil's Rotoscope".

(This stuff, incidentally, is why motion capture is now officially Not Animation according to the rulebook for the Academy Awards.)

Read for more details on all this!


Thanks for the very interesting explanation of why people would make movies using machinima. I haven't seen much of it so far, beyond a few short pieces here and there, but it's a really cool technique for making filmmaking more accessible to a wider diversity of people and ideas, which is a good thing.

Speaking of diversity, there's something that's been troubling me since I read Charlie's blog post introducing you and your work, specifically since I followed the link to I'd really like to hear your thoughts on this.

Would you agree that Machinima would be just as good a movie-making tool for women as for men? I'm only raising this question because on, the tagline "Machinima is a programming movement aimed at young males around the world" [emphasis added] jumped out at me. Prior to that it hadn't even occurred to me to consider it an intrinsically gendered phenomenon.

Is there a good reason why women are being explicitly excluded in that description of machinima? Maybe I'm missing something. In this blog post you define machinima as "filmmaking in a 3D environment rendered in real-time" and "filmmaking using computer games." I don't get "aimed at young males" out of either description. Why is it being phrased that way? Is this something that was on the website when you founded it, or was that added after you left?

Could you comment on the state of diversity in machima in general? It seems to me that machinima, because of the accessibility and low cost of entry that you mention, would tend to facilitate diversity--gender, race, GLBT--compared to more traditional types of filmmaking. Hence I was really surprised (and disturbed) to see gender exclusion front and center on the website.

8:, aka Machinima Inc, is not in any way representing Machinima-the-artform when they say that - they're talking about Machinima-the-well-known-YouTube-Channel.

When they say "programming movement", they're talking about the demographics of the people who watch the shows on their YouTube channel (which are indeed mostly aimed at the young male demographic), and not about Machinima-the-artform at all.

Machinima-the-artform is actually a tremendously diverse movement, with considerably more female directors, in particular, than is average for the film industry. I understand from the guys at the Machinima Expo - who are focused on showcasing the Machinima-the-artform - that they're also seeing entries made using Machinima techniques from all over the world.

I'll ask them if they can comment more on this!

(And just to clarify one more time: I no longer work at Machinima Inc or on the website, and haven't for some years. My involvement with Machinima these days is limited to Machinima-the-artform.)


p.s. I posted about the tagline a couple of days ago on Google Plus; here are some comments people made:

"The video game space is already dominated by males in pretty much every way, shape, and form and stuff like this doesn't help." (Melissa Peterson)

"Um… I make machinima. I know lots of women who do." (Shava Nerad)

"+Shava Nerad That's great you do, but when a well known site says stuff like this it only perpetuates the 'gaming is a boys/mens only club' and women/girls aren't allowed."(Melissa Peterson)

"I watch a lot of Machinima videos. They're pretty entertaining but they do occasionally make sexist comments which make me uneasy. These types of comments are unnecessary and do not contribute to their shows. They probably won't do a thing about it unless they get enough viewer feedback. It would be nice to see female hosts on their channel as well. +Machinima this may be a small post but please take note." (Brent Ching)


Yep, there is a difference between Machinima-the-artform (which Shava is discussing) and Machinima-dot-com-the-website. They are not in any way the same thing.


Thanks a lot for that explanation, Hugh. OK, that makes a lot of sense. It is unfortunate that the website has the same name as the overall filmmaking movement, as it can lead to the assumption that this is being said about Machinima the filmmaking movement in general.

(By the way I posted my second comment (#9) before I saw your reply.)

Thanks again! And I'm glad to hear that machinima in general does have a lot of diversity in it. :)


Off topic for machinima, but of interest to some people here: Fridge Sends Spam Emails.


Re: anime today -- this is a "how long is a piece of string" deal as different Japanese studios do different projects differently and a lot of them outsource the scutwork to Korea and the Koreans outsource the scuttiest of that work to Chinese sweatshops who employ 8-year-olds chained to light tables to churn out the cels. Or so it is rumoured. There was a "Simpsons" episode about this sort of outsourcing work as I recall...

For TV anime of a certain quality standard (Shaft studios etc.) then digital production is the norm i.e. the frames, backgrounds, characters etc. are created digitally using CintiQ screens or similar plus a lot of custom tools. Other shows still use classical cel inking, especially the very-long-running shows like Doraemon (broadcast since the 1970s) where they have a massive library of characters and backgrounds and can assemble a weekly broadcast show out of materials already to hand accompanied by a lot of static frames plus lip-flaps or short repeated sequences.

There is judicious use of CGI and 3D in some places in anime today -- running water effects for example, or fireworks which are now digitally created and look the part rather than being poorly animated if at all during the acetate-and-cel era. Shots can be storyboarded in 3D to get motion and character movement and camera POV moves more correct than 2D animators could usually achieve. This includes slo-mo and even bullet-time effects as seen in FLCL, for example.

There were attempts to use full 3D in anime -- Gerry Anderson was involved in one such project, "Firestorm" which had "UFO"-derived 3D CGI models mixed in with more conventional 2-D animation. Not a success.

Some "auteur" anime creators like Shinkai have gone back to basics, a frame-at-a-time production system which in Shinkai's case in "Garden of Words" was a step back from his historic all-digital early productions like "Hoshi no Koe" which paved the way for the digital revolution in many ways. Miyazaki of course is famous for never having left the painted-cel era.


Hugh, you're an excellent guest poster here, and obviously have big shoes to fill in Charlie. Thank you for breaking down an art form that I admittedly had kind of personally "written off" (at least so far as my personal interest in it is concerned). I'm now going to have to give Machinima ore of a gander, as you've managed to convey what about it that makes it so interesting.

Question: Obviously, video game graphics have continued to improve year after year, and there's no inherent reason I can think of why this will stop any time soon. While we're not there yet, it's entirely possible to imagine that some day not so long from now, video games will be able to render in real time something perhaps akin to, say, Square Enix's Final Fantasy: The Spirits Within (which I love to pieces, despite many people hating it).

What kind of impact do you think this will have on Machinima as an art form? Did you consider this from the early days, perhaps even working on developing the genre with this "end game" in mind?


The only Machinima I've seen in the commercial world is the WOW episode of South Park. Any comments on how well those guys realized the potential of the technology?


I've been waiting for some years now for game studios to release film-maker kits. Given that they have props, characters, mechanics, lighting, rain and dust physics, locations and so on, and presumably they have tools to produce animations for cut scenes, it wouldn't need to be an enormous investment to produce and sell such a tool.

They'd still price it at a pro level, but it would come with permission to use the bundled models, and there'd be an aftermarket for further models, environments, props, and interactions.

Do you think this is something that will happen, or is the market too small for it to pay for the game studios?


@Silly-swordsman - It has actually happened already, a few times over. The best-known (and best) example of the technology is the Source Filmmaker tool, which has already produced some very impressive work: and to see some examples of work made using it.

I'm going to write a post linking people to some resources if they wish to explore Machinima as a technology, either later this month or early next!

@Dave P - The South Park episode was a very entertaining use of the medium. It didn't push the possibilities of Machinima or even WoW, but it wasn't meant to - and what it did do, very well, was evoke the game.

I'll link a few examples of different interesting things that people have done with Machinima when I write the resource post - the range of approaches, styles, and forms of narrative that people have managed to evoke in the Machinima medium is pretty darn impressive.


@nickgold42 - If you look back at the Spirits Within, then compare the stills coming out of the latest Cryengine demos, you'll find that we're actually there already.

I've been doing some tests with graphics techniques that are actually a generation beyond even that - see - and it's fairly clear that the raw visual quality Machinima is capable of is now equal to that of CGI films from 8 years or so ago.

(Because Death Knight Love Story was made over 5 years, it has actually suffered a slight Duke Nukem Forever effect - it's using graphics technology that is at least one generation behind the current cutting edge of games.)

However - and this is a big "however" - there's a lot more to making a fantastic animated movie than just rendering quality. Pixar's rendering tech, in some ways, isn't that clever; likewise, the first Shrek used some shortcuts in places that are startlingly low-technology. What makes them work is the art direction and the quality of the artistry.

And unfortunately, top artists cost top dollar. So the challenge for any Machinima creator looking to challenge Hollywood rendering quality is how they manage to attract and maintain that level of artistic talent without ending up with a Hollywood budget.

(I have a few ideas on that - including "don't bother, go for a non-photorealistic style from the start".)


What makes them work is the art direction and the quality of the artistry.

Indeed so. Pixar's Anglepoise lamp is for me the quintessential example of this. It's a rigid mechanical object with a handful of joints, and yet even the early footage of it playing with a ball show character. With Luxo, John Lasseter truly demonstrated that you can tell story with computer animation.

But then I think Lasseter is, along with Hayao Miyazaki, the best animator working today. I'm amused that one has done so much with computer animation while the other is the best example of painted cell animation, and yet they appreciate each other.


...without ending up with a Hollywood budget.

(I have a few ideas on that - including "don't bother, go for a non-photorealistic style from the start".)

I'd agree; I've no film-making ambitions, but the most immersive experiences I've had have involved coming out of video game cut scene into gameplay, and not noticing immediately because the render quality is absolutely even.

Also, I've had a couple of "not just break the 4th wall, but grind it into powder" moments in films where I've spotted CGI pixelation (no names, no pack drill, but the film was much anticipated, and poorly received by everyone).


Not what I'm into, having cured myself of video games decades ago, but I think you've really tapped into something that will go somewhere, Hugh.


I'm wondering, moving beyond WoW and all that other stuff, is the idea of a generic virtual world, like Hollywood's back lots, in the works or already extant?


There are no virtual worlds specifically for filmmaking (aside from specialised tools like MovieStorm), but yes, there are a few virtual general-purpose virtual worlds which are also often used for filmmaking.

Best-known is Second Life, which has a thriving Machinima community, and hosted the Machinima Expo for many years. Its graphics engine is a bit clunky and it's decidedly ... idiosyncratic in many ways, but it's extremely usable for Machinima.

Currently, the founder of Second Life, Philip Rosedale, is working on a new virtual universe - . Needless to say, I'm keeping a close eye on that particular venture.


Best of luck to in your. Come to think of it, I would see a machinima feature of a Stross adaptation, say, like "A Colder War". I think that's the name of it.


I think Miyazaki is an excellent animator in the sense of storytelling and in the past he was ground-breaking in art direction but the world has learned how to make animation look better on screen over the past few decades (Studio Shaft and Mr. "head tilt" Shinbo, for example). The art of storytelling is something else sadly and less easy to duplicate.

Some of my favourite animes are quite crude in the art sense, sometimes deliberately -- Serial Experiments Lain and the later Habiane Renmei have that signature sketchy art style associated with their art director Yoshitoshi ABe but are quality entertainment because of or in spite of that. On the other hand Shinkai Makoto has been dubbed by some as the successor to Miyazaki, producing quality digital animation attached to excellent storytelling chops. Not much for happy endings though.


About this Entry

This page contains a single entry by Hugh Hancock published on January 23, 2014 3:22 PM.

Introducing ... Death Knight Love Story, evolution of Machinima was the previous entry in this blog.

A Virtual Filmmaking Primer is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.

Search this blog