Final Animation

After compiling the frames in Adobe premiere and setting the audio, I exported the project as a 1080p .mp4 file and uploaded it to Youtube as shown below.

Rendering

The final task inside of Maya was to render out all 275 frames of my animation for compilation in Adobe premiere. To do this I tweaked the MentalRay renderer inside of maya to render a series of frames at a specific using one of the 8 cameras I had positioned and then batch renders the section. For example I would render out frames 1-25 using camera 1 and frames 26-56 using camera 2 etc.

Here is what the folder containing all the rendered frames looks like.

Camera placement

In order to shoot the scene I wanted to have a simple series of cutbacks similar in part to the original source material, it would help maintain the quick pace of the dialogue, so I set a series of cameras 1-8 to capture each significant action and line of dialogue as was laid out in my dope sheet.

Below are two examples of camera angles from cameras 2 and 5 respectively. I tried to go for an over the shoulder kind of angle to keep both of the characters in the shot where possible in order to ground the brief piece of animation and to keep more of the set in the background.

Lip syncing and facial expressions

Throughout the animating process I left the faces and mouths alone for the most part, getting the body language and actions finished first. Then made use of the Max rig's inbuilt facial camera view which positions the viewport camera directly on the face of the rig, allowing the user to manipulate the face. The first step was lip syncing and I made it a priority to complete the basic mouth movements first. I simply opened and closed the mouth to varying degrees in time with the words to get a basic lip sync down.

I then went back once all of the basic mouth movements were complete to add greater detail and definition to the words that were being said, for example pursing the lip when 'o' sounds were being spoken, such as 'do' or 'your'. Aside from this I changed the angle of the lips when certain hard 'T' or 'S' sounds were made.

Lastly I did a final facial pass for each rig altering the facial features in order to convey emotion, such as worry on the privates face when the officer gets annoyed with him, or the anger on the officers face when the privates admits to having no love for his king. Due to the Max rig's fairly simple controls this was merely a series of eye brow and cheek movements to convey the correct degree of each emotion, I also made use of the upper and lower eye lid controls for this purpose.

Animating

When animating my workflow began as thus, I would load the audio file into the frame viewer and listen to it a few times to get a feel for the timing of certain actions in my head, I would also refer to the dope sheet for this. I would then proceed and block out the basic movements for my rigs throughout the animation, making sure that the major key frames were in place. I then continued to animate in passes adding greater and greater flow and detail to each movement, making them more natural and flowing.

I did this firstly for the officer character as he was the 'lead' in this scene, dictating its pace and determining the Private's reactions, so it was easier to animate him first and then animate the private in passes allowing the timing of the officers actions to dictate the privates reactions. Aside from order, the animating process for both characters was much the same and it involved referencing many time to the reference material collected, especially on how the arms moved when the officer would bring his arms out from behind his back or for how his feet would move when he turns around.

Locators and props

One of the hurdles I had to overcome when animating was keeping the props, namely that hats, glasses and mustache, attached to the characters when their heads moved. To achieve this I used a system by which I created a series of locators and parented them to a specific control on the rig, namely the head control. I then parented the locator to the prop I wanted to remain fixed.

Secondly I wanted to make the private character at one point reach behind him and reveal a picture of the king, to do this I locator bound a modeled frame with a photo texture to his hand and then hid the model, creating a key frame at which I would re toggle its display options, making it visible again. This gives the illusion that the frame is not in his hand until he reaches behind him, when it then appears.

Scene reference

As a setting for my scene I decided to create a room styled after a WW1 dugout, I took a look at some images of dugouts and selected a few prominent items that appeared in each. These ranged from bunk beds to book shelves but it still made for a convincing setting.

I also made use of the crate and spellbook models from my texturing module for some amusing cameos.