#milestone1c

Milestone 1C

The process of producing this weeks milestone involved many techniques that were known but had not been used in a long time, many of which had been unused for long enough that they needed to be re-learned.

The first step involved updating the environment. This was done in blender due to the familiarity with the program and was relatively simple to complete. The models were divided down into more separate pieces, and each piece was separated into multiple materials based on the concept art.

The environmental models did not require much in the way of learning, and mainly consisted of re-hashing and reinforcing existing knowledge.

After this, a new character and animation was envisioned: an overweight and incompetent and ungainly opposite to the example provided for the milestone video.

Unfortunately Art v1 had decided to stop working; attempting to move past the second stage of the rig creation would simply not work, no propmts would appear and no way to continue on in the process was available.

Even attempting to use the computers in at university proved to not be viable, as the rigs brought home raised errors on opening and could not be used.

Finally, after perhaps a total of 4 hours spread across multiple days and multiple reinstalls of all the involved software, it was decided that the older prototype model from milestone 1B would be re-used with a new animation.

The process of animating itself was no different from the last milestone, though a number of points that could be improved upon were identified.

using reference footage, for one, could have improved the quality of the animation, and in the end the action was too subtl, and could do with a fair amount of exaggeration to more effctively portray the character.

Bringing the new models and animations into Unreal and modifying the animation speed in the sequencer provided no new challenges, though a number of experiments with lighting and fog were made.

First, changes to the sky-light and the directional light in the scene were made in addition to exponential height fog to create a more interesting environment.

After these were done the character was provided, first, the default epic games texture, and then later a customised texture with subsurface data to create a more skin-like appearance.

Once the new animation, lighting, fog and environment were all implemented to satisfaction, the sequence was animated out into 3 seperate passes: final image, depth and subsurface colour.

These three sequences were imported into after effects and, like before, provided no new challenges due to the similarity to milestone 1A.

The subsurface color was tested and then rejected as a color mask, due to the even sharper edges it created, and in the end the depth pass was used as a luminosity mask due to the more interesting effect. Further, the video was modified to have higher contrast and modified levels to counter the faded appearance that the fog from unreal created.

Time in separate programs:

This week the main thing topic I had known in was definitely the animation of the character, but I feel as though much like last week, I rushed it a little. This was because my Mac at home couldn't operate the ARTV1 toolkit on Maya. So I had to stay back at uni to complete the Maya section of the Milestone. This is probably the where I spent the most time though, after creating a character that I was happy with (which ended up being a lot easier this week as I knew the process,) I had to figure out the actions I wanted the character to do. I ended up going with a shotting taunt as you can see from the video. Then animating the character proved to be more difficult than I intended it to be. This was because I didn't have the clearest image of what I wanted the character to be doing in my head. To streamline this process and effectively get a better quality animation, I should have drawn thumbnail sketches before hand to keep me on track.

The next thing I did was conduct the lighting experiments on the still character. I have talked about why I chose each style of lighting in earlier posts, but after experimenting, the final style I created was a mixture of the two. I realised that with the Rembrandt style image, It would be hard to see the whole animation play out and I wouldn't be able to create the effects as effectively in the final video as I did. But I wanted the side angle style that the top down spotlight didn't offer. So by having the light angled like the Rembrandt, but increasing the cone angles, I created a softer light that lights more of the scene. If I had more time on this milestone, I would experiment on more different and creative lighting conditions. The lighting of a scene is something I haven't had much experience in but it fascinates me how the lighting itself can tell a story.

The final part of this production was to compile render the camera movement and compile it in After Effects. This part of the production was one of the easiest parts of production as I have had a lot of experience in it. I put basic effects on the hue and saturation of the video that animates when the character shoots his gun hand. This was a simple effect that gave a lot of emphasis on the action which is something I wanted to achieve.

5-6 hours production

Here are my two different lighting environments, The first is a night time environment with some harsh lights bouncing off the character

here is the final creation unreal crossover with after effects, as previously stated I used a form of green screening with a alpha luma matte to achieve the text behind the polished footage. I had a lot of fun creating the character with the ART maya tools. I just experimented until I achieved this ape/human creature.