With the third season of The Walking Dead behind us, Stargate Studios have released a video showing the dismembering of special effects for the season.
Note: The video mentioned in this article can’t be embedded for some reason, so I’m just going to link it instead. Enjoy it here
AMC’s beloved zombie show The Walking Dead has just ended its third season, and thus it’s time to check out some behind the scenes footage of how the special effects were created for the show. You may be surprised to learn that most of the zombies (or walkers, as the protagonists call them) are simply wearing very elaborate makeup, and plenty of the skull crushing and zombie smashing is actually done with blood- and gore-filled puppets. Still, there are times when impressive on-set effects won’t do the trick, like when you need to bury a machete into someone’s skull. For those times, there are video effects, or VFX.
Amazingly, this isn't actually done by computers
Adding VFX to a video is often an elaborate and tedious job, and since the video isn’t narrated, I’ll give you a brief overview instead. Special effects are layered on top of video footage, much like transparent celluloid with drawings of Donald Duck were layered on top of background drawings to create Disney cartoons in the past. Each layer contains some new element that is added to the scene; this can vary from a filter that alters the color of the video, to smoke, bullet holes, fire and entire buildings or backdrops. Most layers containing objects to be inserted into the scene, are filmed beforehand on green-screen or with a dark background, so that they can be cookie-cut into the right size and placed in the shot.
The video then needs to be tracked. In simple cases, this means focusing on a tracking object in the shot and observing how it moves, rotates and rescales across the screen as the camera moves. Simple two dimensional effect layers can then be added to the scene, and blend in by copying the same movements as the tracking object did.
Adding this computer modeled building to the scene requires more than simply tracking an object to ensure it blends in.
If more complex layers need to be added, such as 3D-modeled buildings that need to be seen from many angles, the process becomes more complex: While layering works the same way, the computer has to be told to rotate the computer model. For example, if a camera rises from ground level and up 30 feet, our 3D-modeled building has to rotate so that we initially see the walls, but then see the roof as we rise above it. A simple model of the scene is built in the video editor to map out where the camera is in relation to the scene, and then the camera, as opposed to the footage, is used for tracking. Similarly, adding features to a zombie, like a computer model of a missing jaw for the character, would involve modeling the actor (as opposed to the set), and using that model as a reference.
The number of effects layered onto the video can become staggeringly large. In the effects breakdown, there’s a short sequence where an assault rifle is fired into the ground and the bullets knock dust into the air wherever they impact. In all likelihood, every one of those dust shots is added separately, colored to blend in, blurred if the camera moves and cropped to make sure they fit in nicely. The same goes for the shell casings, though it’s likely not every single casing is separately animated.