Week 4 – Tests and Experiments

This week I plan to iron out any uncertainties and make experiments in the 3D techniques that will be used in the assets creation stage. By assets I mean each animations that I will complete, I consider each of the as a game/art asset. As I mentioned in last weeks post, I had a discussion with my tutor about how to do certain things that I was unsure of when my group discussed the animations I would need to provide. Hereafter is the extent of my preparation to complete the assets. Each asset has been assigned to be complete within about a weeks period. Therefore, after this testing and experimenting week, I should have all of the necessary preparations to complete the animation asset smoothly. At the least it will give me a head start and introduction into areas that I am not knowledgeable about.

Animation Tests

Bee Wing movement

I have certain options in the way that I can animation the bee’s wings flying:

• Simple up-down motion (flat, oval shaped plane).
• Curved at extreme poses and straight/level at the horizontal.
• Look at this reference of bees flying and make the shape and movement more realistic.

I want to first make some quick tests of all of these options and see which look the best, and at what speeds give certain styles e.g. cartoony or realistic.

I initially wanted the wings to have the visual effect of the image below, focussing only on the wing appearance not the bee cartoon. I am not sure how to do this with a 3D model, but with motion blur added in Unity and a sufficient wing speed, they should look like this.

Bee Wing Animation Test/ Experiments to Try

Play blast Wing Experiments – Simple Movements

Here I tested the wind frame speed and wing angle, as I stated in the diagram above. (Just to clarify, the bee models are not the final models, only for the tests). The frame wing flap speed can be seen in the above video, all those in the first column of bees. From the bee up front to the back, the wing flaps every 10 frames, then 5, 4, 3, 2, and 1. The other two bees along the row show a diagonal and horizontal movement. I am yet to decide the angle of the bee wing movement until I finish the more complex rig wing movement test. Nevertheless, the frame speed can be analysed now. The wings that move every 2 frames had the best result in my opinion. It showed the motion blur effect that I aimed for well. Where as the wings that moved every single frame didn’t have the same effect. It was as though you could only see the wings at the extreme positions and not elsewhere. As though there was no motion blur.

The best

Wing Experiments – Complex Movement

As in the diagram above, I wish to experiment with a more complex wing movement. That incorporates a rig movement with rotation and translation, instead of the simple movements above. I found a section of the bee wing movement reference above (Youtube video), which best portrayed the type of movement I wished to create. Therefore, I used it as a reference for the next rig, perhaps even rotoscoping it.

I found a reference for making a rig.

Cloth Test

The idea that I have, for the cloth blowing in the wind animation, at the moment, consists of a heavy feeling material. The wind will be strong but the fabric will be quite thick and heavy, like old canvas, so it doesn’t move as much as e.g. silk may do in heavy wind. I really like the ribbon like cloth in the game ‘Journey’. So I would like to experiment with types of fabric, and shape, placement in scene.

Using N-cloth and point constraints.

How to point constraint: Select vertices on plane>Transform constraint.

I have come across many problems in the past when using n-cloth. Such as the stretching of the mesh at the constraint points (see image below). I have a few methods to fix this, although, it is a hassle.

I found another video that really helped. I had always used constraints on vertices, but using them on faces seems to work better. There are less stretchy glitches.

I decided that my cloth tests right now (to prepare for the actual animation), should focus on the wind speed, wind noise, gravity and type of cloth (using presents). I chose leather first, but others had some interesting effects also.

The playback speed was incredibly slow, I though this might be a computer or software issue. However, after I decreased the geometry on the cloth planes, this fixed the issue and my workflow increased speed.

When testing out cloth shapes, each one had a different response to certain circumstances.

There was also a method that I discussed with the group. To help with the placement of the cloth in the Unity scenery, I wanted make the appearance of having ripped cloth tied to a tree. However, they would be done with separate models, perhaps with separate animations. One would look like cloth circling a tree, the part of the cloth that is tied to the tree. The other would be free flowing fabric that has constrains on one side (the part of the fabric that is free flowing in the wind).

the games designers would have the freedom to rescale and place each one of theses on any object they wanted, and choose based on suitability. I could also make a few different versions, in different shapes in order to produce variety and give the games designers choice while they work to fit the animations to the scene.

With the ‘around the tree (or other object) cloth animation, I believe that the preset ‘burlap’ is the best. It doesn’t move too much but you still see enough movement. I can apply this animation without needing too many constraints inside the cloth model. Where as if I applied a softer material preset such as ‘t-shirt’, I noticed that it changes its shape too much, even going through the tree. By the way, I tried to add a passive rigid body to the tree (so that the the cloth does not pass through), but it didn’t work.

I believe this is because I have already created a gravity with my n-cloths. So, that I wouldn’t waster time or the energy I spent on my nucleus settings, I decided to leave this for the testing phase. It is not important.

Cloth Play Blast Setting and Shape Tests

Play Blast Shape Analysis

The first rectangular cloth had nice movement, although it alwasy seemed restricted.

The long ribbon/flag shape had much manuverability and twisted a lot. Not sure if I like the style of movement

The cone shape hanging off of the tree had irregular and not aesthetically pleasing movement.

After experimenting, certain times the material more or less responsive to the nucleus settings. When I added more geometry to the model, it didn’t move as much.

Quadruped Rig

The test for this asset will have to wait until the assigned week for this asset. Since, I ran out of time for the experimentation week.