Making stuff move
Some of the beady eyed amongst you may have noticed that there isn’t much animation in my blog. So, I am going to (gradually) right that wrong and explain – for the uninitiated – what it is that I do (apart, it goes without saying, writing tracts of entertaining prose for you on my blog…).
I animate in both 2D and 3D. For those of you who don’t know the difference: cut out a picture of a brick and hit yourself over the head with it. That’s 2D. Next cut out six pictures of a brick taken from all six sides, stick them onto six sides of a cardboard box and hit yourself over the head with it. It may hurt a bit. That’s 3D. Now get a real brick and hit yourself over the head with it. That’s stupid. But would be loosely classed as stop motion. I don’t do stop motion.(Some examples of each discipline are: Snow White: 2D. Toy Story: 3D. Wallace and Gromit: Stop motion).
I will take you, then, through a short and simplified account of how 3D is made. This is neither the way it should be done or shouldn’t, it is just the way that I do it (before you 3D dweebs start scuttling out from under whatever rocks you skulk under – give me a break, okay?).
- Storyboard / Script This is out of my hands most of the time as the script or storyboard has been already written / drawn beforehand and I have to follow it the best I can. There is obviously scope for me to say that certain things are not possible in the timeframe or may not work visually, but I have to crack on most of the time as I am on a daily rate and clients hear the clock ticking over all other ambient noise. I use this time to sit down and plan what I need to make, what order I need to do it in and how I’m going to do it. This sometimes involves panicking.
- Modelling Thankfully this is not when I whip my kit off and stand in front of 20 gormless art students (I was a gormless art student once…). No. This is when I make, in 3D space, an object. It’s complicated, and one day I’ll devote an entire post to it, but in essence it’s this: you make objects out of things called polygons (and before you 3D nerds raise your misshapen heads: I know, you can use Nurbs or triangles – but this is complicated enough). They have four sides, otherwise there’d be holes when you stick them all together and that’s no good. Try to think of them as square pieces of card that can only be stuck together at their edges to make an object. Like a cube would have 6 squares but a sphere would have 64 squares, say. When I make a character or a complicated object I may have thousands of these squares.
- Texturing & Materials Once I have finished my object I need to give it a colour. If I made a milk carton then I would need to make all the polygons white and then put a picture of a cow on it. This is called texturing. The way that you do it is to unwrap the object (like a flat pack) and then you can paint on the ‘flat pack’ in Photoshop and put your cow on it. This is called UV mapping. You can also make your object shiny, reflective or bumpy.
- Rigging After making my object I will want to make it move, to animate it – unless I am making an experimental Warholesque film about still objects. If it is just a ball, then I can just move it around in 3D space. But if it is a character then it’s obviously a bit more complicated. I can’t move every polygon, so I need to use some tools to move large groups of them. These tools are called deformers. A rig is an animation term used to describe a system of controls created by the animator to move the deformers. I don’t like rigging; it’s boring, a bit complicated and ever so slightly geeky.
- Animation This is the meat and potatoes of what I get paid for. The basic principle is that I tell the computer that an object is at coordinate A at 0.00 seconds and then at coordinate B at 5.00 seconds and the computer moves that object between the two coordinates over 5.00 seconds. Capiche? It sounds so simple when I write it down…
- Lighting Did I say I had to also be my own gaffer? But instead of having a big Arri truck parked outside my gaff every project I can ‘merely’ plop virtual lights down anywhere I want. Lighting is something that I find quite challenging. The problem with lighting in 3D is that you can have as many lights as you want so I end up literally painting with light…which cause problems with shadows and, ultimately, rendering times…which brings me to:
- Rendering Once I have made something, textured it, rigged it, lit it and animated it (yeah, I am the film crew) the computer needs to ‘paint’ each picture of the animation (every second, 25 images have to be made). Depending upon how complicated the animation is it can take ages. What 3D studios have is something called a render farm, which is a bunch of computers (know in the business as nodes) that are all linked together that render quicker. For example, if I made a sequence that takes 5mins to render each frame and my sequence is 10secs long then that will take nearly 21 hours (250 x 5mins / 60) but if you have a render farm with 250 nodes then it’s take 5mins…depending upon the speed of the computer. Render times have a massive impact upon my work. I want a render farm for Christmas. Really.
Having just read back on this, I understand that I am going to lose a large portion of my ‘readership’, but possibly gain an army of nerds with bad clothes and hygiene issues.
Oh, and in terms of kit (for anyone who is interested) I use an iMac Intel (maxed out GFX card and RAM), Maya 2008 Complete, Z-Brush 3 on XP in bootcamp and Photoshop CS3. That’s it.