Group Meeting and Notes:
During this meeting, we met the sound design students who would be helping develop the soundscape and effects for the game.
Personal Goals:
. Set up the scenes for the interactions
.Animate cycles
Scene References and Organisation
While beginning the scene set-up process in order to effectively organise our scenes, I encountered several issues which created some major setbacks in our progression of animation. When moving the global control of the seagull, the main body mesh separated itself and offset from the main rig controllers and the eyes. This issue seemed to occur in both rigs, so looking into solutions for both rigs I found that the root issue was in the rig’s entire hierarchal structure, which needed to be edited to make more coherent sense.
With a lack of understanding of the rigging process, when opening the referenced file to investigate the issue at hand, I tried to delete the history which deleted several elements that destroyed aspects of the rig such as the Wing Iks. As stated by Watkins in his book Creating Games with Unity and Maya: How to Develop Fun and Marketable 3D Games (2012), He expands on the importance of deleting its mesh history before beginning the rigging process to minimise the file size and prevent and excess of nodes, so critically reflecting on this in future, our process of rigging will have to be much cleaner. In this light, I unbound the rig from the skin in order to reset the mesh and deleted its history as I found several issues to tackle.
Rigging Issues and Solutions
One of the key elements we had to consider when improving the rig was the addition of IK controllers. When going to create a test walk cycle with our characters to provide to the VR students, I noticed that with a lack of IK controls in the legs, full-body animation would prove excessively difficult for us in the long run. Taking this into consideration, I research and followed some YouTube tutorials in order to gain an understanding of how leg hierarchies work in an IK Format.
Following this, I created a test that displayed how this worked in order for my fellow teammates to see and understand how I intended to adjust the lower half of both the rigs to improve the mobility and fluidity of the animation.
Seagull A
Applying this to the first rig, Seagull A, in which the legs are longer and more replicant of real human legs, the implementation was fairly smooth and easy to calculate. The real difficulty in terms of achieving ‘realism’ or anthropomorphised Seagull behaviour will be in the skin weighting and directional knee bends.
Lining up the ‘knee’ controllers and placing the skeleton IK into the seagull in order to correctly place the joints and attach the new IK legs to the rest of the rig. Understanding joint connection hierarchies and IK applications, a key element I learned that can affect joint rotation is the ‘Rotate-Plane Solver’ which changes the way that the joint can follow a pole-vector control.
While it is not vital to gaining directional control, it is possible to set the IK directions early in the rigging process but bending the ‘knee’ joint towards the direction of intended influence, which helped my understanding of knee and pole vector rigging practised to make them smoother in the future.
Once the rig was successfully connected in a sensical hierarchal order, the next key step was to create a pole vector constrain and control that could control the knee pivot. This is a key element to rigging with IK controllers, that in retrospect could have been used to create directional control for the wings, and is something I will heavily consider in future rig creations, especially in more animalistic bipedal rigs.
A learning step for me was understanding how to use and implement set driven keys. Going back to the reference of the seagull movements, their feet roll in a similar fashion to humans; typically in CG rigs, extra features are created in the attribute editor for the ease and access of animations to create the most realistic animation possible without breaking elements of the rig itself. So In creating different controllers for the foot, ball, and individual toe joints, I was able to add set driven keys to that group to create foot and toe rolls that can be manipulated in the attribute editor. Following the tutorial from earlier, the process of creating toe and foot rolls was an extremely useful tool that would help myself and Marianna key aspects of the rig in a more clear and less ‘messy’ way without direct animation onto the joints of the character, which cannot be frozen in transformation.
Expanding onto clean rigging pipelines, I thought it may be worthwhile to limit the translation and rotation information in order to effectively restrict the motions available to the character so they make logical sense to the world. For example, in limiting the X-axis information in alls the rig to stop and be more proportional without excessively stretching and contorting the joints and therefore breaking the mesh of the skin. This will help to build a level of consistency during the animation process between me and Marianna as the ‘volume’ and shape of the character will appear the same, and also add to its character believability.
In addition to the IK leg control, I also found issues with the way the eye controls were functioning on the seagull character, as they were directly parented to the eyes rather than aim constrained, which is a much neater method of rigging eyes as it does not directly affect the translation of the sphere, but instead only its rotation.
Below showcases all the rigging adjustments I was able to make on the Seagull A rig, including IK controls and eye rigging controls.
Eye Rigging References
Rigging for Beginners: Eyes and More in Maya – YouTube
Skin Weight Paint Influence Issues
One of the initial problems I went in to fix was the skin weight issues on seagull A, as both feet had influence over one another, so they could not be lifting without dragging the polygons of the other along with it. Experimenting with different aspects of this, my first attempt was to add a negative paint influence to the faces affected on the opposite foot, however, despite my best efforts in this, the influence seemed to not remove itself. After attempting to find several solutions to this, I discovered that the method that proved most effective in this case was flooding the affected vertexes with either 100% Positive or 0% negative influence in order to get them to act more realistic in motion. While this worked and I was able to more efficiently paint the weight into even and realistic weight distribution, this seemed to point out several issues in the mesh hierarchy, as in flooding these vertexes the wings were also entirely affected and followed the feet rather than the main body mesh. This caused further setbacks in the rigging process but offered suit for some critical reflection on rigging pipelines for myself. In reference to Watkins, he states how Maya in the process of skin binding will attach its vertices to the nearest joints on the initial bind without the contextual physical and biological context we have, and since the wings were not attached to the mesh via ‘combine mesh’ attributes, the vertices must have attached themselves to the feet without spacial context ( 2012, p. 303).
Seagull B Rigging Issues and Corrections
Taking the same developments from my rigging progression of Seagull A, I applied these to the seagull Brig. Since learning the process with seagull A, I was able to apply these changes faster with more accuracy, which led to a cleaner overall rig.
As seen in the video above, the older rigged version of the eyes was not very effective for the animation process, so adjusting and changing them to aim constraints gave us much better control over eye rotation and look much more realistic. In order to maintain and preserve the style and character of the rig, instead of creating an aim constraint that moved both eyes at the same time, I added controls that matched the distance offset between the eyes so they are controlled separately, to upkeep the confused and ‘stupid’ look of the bird.
References
Adam Watkins (2012) Creating Games with Unity and Maya : How to Develop Fun and Marketable 3D Games. Burlington, MA: Routledge. Available at: https://search-ebscohost-com.arts.idm.oclc.org/login.aspx?direct=true&db=nlebk&AN=376905&site=ehost-live&scope=site (Accessed: 5 April 2022).