I wrote this brief overview on computer graphics for a presentation I did last month.
Over the years I’ve found that there’s a lot of people that seem interested in how computer graphics and more precisly movies like Toy Story are made.
So I decided it might be of interest to some of you as well. It’s only a very rough overview and obviously there’s a hell of a lot more to the art of computer graphics but hopefully it will give you some idea of what goes into making these amazing movies.
I created this image in 2005. The face was based off of my Grandad.
CG stands for Computer graphics
After the script has been finalised just like with traditional movies, a CG movie should start with a story board.
Once a storyboard has been drawn, animated movies usually have what’s called an animatic or Pre-vis. This is basically a moving storyboard. It gives everyone a much better sense to what the final movie will be like.
This link below shows the rough version of the short animation I created Baggage. This started off as very basic blocked out animation (Animatic) As I completed the animation I’d replace the rough version shots with the completed animations. In the video you’ll see that some of the animations are still very basic, these are from the animatic.
You’ll also notice that none of the shots look very pretty, this is because they still need to through a process called rendering which I cover further down in this article.
Once everyone is happy with how the story flows production can start.
Unlike live action films in CG you don’t get anything for free, everything has to be made from scratch. From huge skyscrapers right down to plug sockets on a wall. If it needs to be in the film then it has to be made.
Although just like sets on movies, you can cheat and make facades. Below is a link to another blog post I did a while back on the creation of the street scene also featured in my short Baggage.
Making a CG street
Most computer graphic models are made up of shapes called Polygons, These are basically flat squares. The more polygons you have the more detail you can have in your models. But with more polygons the longer it takes the computer to process. It’s always wise to use real world reference to aid in the building of your models but you can also use photos as a basis to texture (colour) your model
Once you have built your models you need to add colour and materials to make them look real. These can be taken from photos or created using algorithms within the computer. The term for this process is called texturing and shading. Texturing refers to the colours while shading refers to how light will interact with the material. Just like in the real world, where light bounces of materials in different ways, the same can be achieved in the computer. For example think about the difference between a tennis ball and a pool ball.
Pool balls are shiny and reflective
Where as tennis balls have no reflections and thus no shine either.
In this following video I take a simple brick texture and apply it to a surface. Using separate black and white images I can tell the computer how shiny the object it is as well as how bumpy it is. These separate images are usually replicas of the original colour image but have various grey scale intensities that allow the computer to know how shiny or bumpy you want your surface to be.
NOTE: Please excuse the “Demo Mode” across the middle. I was using a capture software that I downloaded so they stick that across in till you decide to pay for it.
Texture shading demonstration
If the object being created is a character or something that needs to deform in an organic way, then it needs to have bones, very much like we do. This allows the object to bend.
Where the bones are placed dictates how the object can bend. And just like a puppet these characters need to have controls so that the puppeteer/animator can control the character. This process is called rigging.
Once the character has been rigged it can be given to the animator to animate.
There’s a common misconception with computer animation. A lot of people I speak to believe that computer animation takes a lot less time than hand drawn animation. The reality is that although the computer does make some things easier, there are others parts that take longer and so both methods take a similar amount of time.
In the link below I show a brief demonstration of how an animator would use the controls on the characters to put them into a pose. Again please excuse the big “Demo Mode” across the middle.
Posing a character
The animator will of been given a shot to animate. They sometimes get the animatic as a template to where the character needs to move to and how long it should take. If there are a number of characters in the one shot it’s usually the responsibility of the one animator to animate all of the characters.
Usually an animator will shoot live action reference of them acting out the actions that the character needs to do. They may also sketch some thumbnails to get an idea of how to add appeal to the poses.
They then need to create all the main key poses that express the movements and feelings of the character. Once they’ve done that they will add in all of the “in-betweens” which is the part of the action between the main poses. Below is link to a shot from the short Baggage where Sam pulls his bag from some Tube train doors. This stage is called blocking which is where all the main poses of the animation have been put in place.
It’s also worth noting how some of the poses I’ve stretched his body more then would usually happen. This is a principle of animation called Squash and Stretch. There are 12 key principles in animation which were thought up originally by the nine old men at Disney. You can find out more information about these principles here Principles
The next part of the process is to pass the animation onto the rendering and lighting team. This is where the sets and characters are lit and finally a process of what is called rendering.
Rendering is where the computer takes all of the information that has been created. So things like where the polygons sit in the virtual world, what moves where, how the lights affect the shading and textures, what’s solid what’s transparent, what’s reflective. All this information is computed so that we the viewer get to see the final result.
And that is a brief overview of what goes into computer graphics.Like I mentioned at the start this is a very basic overview, if you’re interested in learning more then give me a shout I’m more than happy to talk about it or point you to some good websites.
In till next time!