Character Movement

These are some of the videos we looked at when considering how our character could move.

The ink creature in Sang d’Encre (2012):

The water fairies in Maleficent (2014):

Mermaid Swimming:

We really liked how flowy the ink creature was and I thought for a while that we could do something similar with nCloth.. Clare found this tutorial on how to create wings with nCloth:

When making Anam’s head scales I tried to model them similar to wings with a skeleton on each side and a plane stretched between. I had planned to make them look like they were caught in an ethereal breeze but I didn’t get enough time to figure out how to make separate nuclei to act on the cloak and head scales (I thought that maybe the head scales needed less gravity but there’re probably other ways of going about it).

Advertisements

Rendered Head and Topology

Okay so i might have went a little over saturated with the purple lol. The lighting also shows up the flaws in the topology, like I think I may have gotten the flow wrong over the muscles in the neck.

Andrew Loomis – Drawing the Head and Hands

These pages from Loomis’ book Drawing the Head and Hands was good for remembering how the facial muscles flow for planning topology.

Siobhan’s Head Topology

This is the mesh that I drew on top of Siobhan’s face for reference during modeling. I had also aligned the front and side views in Photoshop but there was still some areas like the nostrils that didn’t line up perfectly in Maya.

This topology/facial articulation reference from hippydrome.com came in very useful:

step4a

Head Modeling Progress

These are screenshots I took as I modeled Siobhan’s head. I mostly used three references for this topology project:

Project 2: https://courses.cs.washington.edu/courses/cse458/14au/content/projects/project2/

Hippydrome: http://www.hippydrome.com/

Thundercloud: http://www.thundercloud-studio.com/index.php?page=tutorial/ModelingTutorial/headModeling

I started by drawing Siobhan’s side profile with the quad draw tool. I similarly drew quads around the eye shape from the front view and then connected these to the nose. I initially did this with the quad draw but then found that extruding edges and using the target weld tool was easier for bridging connections.

I drew the mouth with the quad draw tool and extruded the edges a few times to create the loop flow around the mouth. I looked at the reference from hippydrome.com a lot to keep in mind where the poles occur at certain areas where the eye, mouth and nose loops border other eachother. I mostly used the front and side orthographic views when shaping the mesh to the images of Siobhan.

I then started to bring the geometry from the front of the face around to the back of the head. I first made the mistake of bring the geometry straight up from the forehead resulting in too much resolution being directed to a small area. I looked at the Project 2 tutorial again and found that the geometry needs to be spread around the temples also. I made a similar mistake with the edge flow leading down to the neck and had to rethink how the loops from the jaw and neck flow back around the face.

The sculpt geometry tool came in useful from this point on. The smooth function helped tidy the mesh while the push and pull functions helped define the varying volumes better. I then created Siobhan’s ear by drawing around her ear shape with the quad draw tool. I tried to keep the resolution as low as possible so that it would connect to the mesh of her face without creating n-gons.

After this I tried to reduce some of the resolution leading down the neck and top of the head but it had to fixed when I mirrored the geometry as it created poles.

For the eyes I followed this tutorial that Abigail showed me on Youtube:

I pulled the cornea a bit more out at the front to match the diagram on the Thunder Cloud tutorial. I also added thickness to the eyelids and added the canthus. After the mesh was mirrored over I spent some time on adjusting the asymmetrical details.

Lastly I used the ‘mesh cleanup’ to find n-gons and triangles on my model. I cleared some up around the ears and nose but left an n-gon on each side of the nose and on each ear.

Lighting and Texturing Slán Plus Revisions from Feedback

Scene 01:

We were told to bring back the contoured lighting that we’d painted in our greyscale animatic. For the first scene I tried to make the light catch the curve of the hill more but I didn’t want to bring too strong of a light into the foreground as the sun was low in the sky and the darks in the foreground created a feeling of depth. I had tried sculpting a much earlier version of the hillside with light over Easter but it didn’t work very well (plus my hill model needed some smoothing):

hill_02_render_01_hd

trees_hill_render_01

It also became apparent from these tests that if too strong a light was shining on the trees that their colours could unbalance the scene quite easily. Composing the different assets together was difficult and I was worried that everything was looking flat. I had started to make different trees that looked more wind blown and pointed within the composition (like in the greyscale animatic) but then Clare managed to compose the assets in such a way that they didn’t look half as bad. These were my lighting tests (ignore the edge of the sea lol):

My trees and clouds made this scene harder to work with as all the files in the layered shader seem to increase render time a lot. Alec showed me how to bake the separate tree textures in render layers which might have made it faster. However when I tried it a second time the alpha refused to bake even though I had the setting on:( It’s something I’d like to look more into as the long render times makes it a lot more difficult to test different lighting setups and tweaks. Having a layered shader also complicated things for Mark when it came to making Anam disappear at the end so that didn’t happen. This is the revised lighting for scene 01:

slan_revised_scene_01_.0151

The second years pointed out that the sky was looking very stretched in some of the scenes. The stars also became too large in some instances as they were painted directly onto the background and we had to scale the sky so that the camera wouldn’t catch the edges. I made a longer rectangle plane, brought it’s UVs into Photoshop and duplicated Abigail’s sky across and made the stars smaller on a separate layer so that we wouldn’t have to scale and stretch it as much. However this revealed more of the colours in Abigail’s sky so that the palette is slightly changed in our revised version and there’s more detail in the sky. Haha it really is reminiscent of the aurora borealis. I worry that we’ve not limited our palette enough but at the same time we have managed to create a fantasy feeling that we wanted.

The texture that I painted for the soil rendered very pixelated. I think it’s because the camera is zoomed in very close to a small area in order to frame the potatoes. I ended up using a Maya mountain texture instead.

We also had to fix the grass planes in this shot as when I duplicated Abigail’s grass I neglected to separate them in the z-axis which caused them to flicker in the animation (at least I think that’s why).

Scene 02

For scene 02 we were quite happy with the contoured lighting that we managed to get. This is the texture for the cliff that I prepared (the different textures are saved as separate files with alpha for masking in Maya). Haha the grass brush looks a little silly, like an Hawaiian printed shirt so I didn’t make it stand out. I added paths to the cliff to hopefully lead the eye and suggest others have travelled this way. I also added gradients to the front of the cliff to hopefully make it appear that it’s catching the light more.

Abigail then set up the rest of the lights and spent some time making sure that the shadows of the rocks weren’t too black (bounced light from the sky). She also did a good job of making sure that Clare’s boat had some atmospheric depth affecting it, making it appear further away and bigger by adding a lighter valued shader. James suggested that our cliff was looking very smooth so I went back and bump mapped one of the textures.

slan_scene2.0001_lightened

In our first rendered animation the particles that Mark set up unexpectedly rendered black. Clare tried turning all the lights off and noticed that they rendered brightly again. I looked it up and found in Maya’s user guide that if you render a scene without lighting Maya will create a directional light that’s parented to the camera so this must have been why the particles worked in Mark’s test renders and not the main scene. I shone a directional light onto the particles and light linked it onto them only which seems to work fine. Then the brightness of the particles can be controlled by the intensity of the light.

The first time around we forgot to add the clouds to this scene which made it discontinuous with the next scene. I like it without the clouds too but for the sake of continuity….oops. This is the revised version with the extended sky and clouds added:

Scene 03 (or scene 02 shot 02)

For scene 03 we noticed that Anam’s silhouette was getting lost a little against the background, especially as the sun set. Mark had tried picking her out with light but I think the values and textures in the sky maybe overpowered her a bit….?

Mike suggested that we include part of the monument in this shot to give it more context. Clare set this up and Abigail added more texture and light to the underside of the monument. We had discussed adding stronger rim lighting to her ears and face but placing the monument behind her made her pop more anyway. Clare also added more ships to the composition on Mike’s suggestion to show the large scale emigration.

Scene 04 (or scene 02 shot 03)

For the last scene I think I didn’t give enough time to go back in and darken the cliff’s light gradient that worked for the second shot. I think the contrast was weird as a result and maybe hindered lighting the forms for Abigail the first time around. The stars were very apparently scaled too much in this scene also. It might have worked but Mike pointed out the big one as being distracting.

For the revision I darkened the cliff texture and tried to make the cliff and monument forms feel more 3D with light. Abigail also suggested that I add some more detail to front of the cliff.

Mark also tweaked my lights a little to make it more night time like and removed a light that was too strong on Anam for the new closer camera shots.

After Mark added the new animations to these scenes I also had to go back in and make new cached nCloth simulations for the cloak. This was going okay until I got to the last scene where the cloth simulation wouldn’t play back at all. This was very frustrating but it eventually worked when I imported Anam and a new cloak into a scene by themselves, cached the simulation alone and then re-imported it into the scene. Mark then added the final particles and I don’t think we were ever so happy to press render on Friday night/Saturday morning. Abigail and I also learned the pain of not saving often enough and loosing work that night. I think we were just too tired lol.

UV Layouts

These are the UVs I unwrapped (with a combination of cylindrical and planar mapping) and laid out for Abigail’s rocks and monument models so that she could texture them.

Alec mentioned that for next time it would be simpler to unwrap one rock, duplicate it and then modify it slightly so that the UVs would be almost fully unfolded and then repeat for the other rocks instead of starting from scratch for every rock. I also saw on digital tutors that you can transfer UVs between objects but it wasn’t working for me when I tried it on the rocks.