FMP Research
This page covers my research for developing the Final Major Project.
Please click the rounded images for the link to the site.
Last Term Transferable Research
Through the image below, a link to my work from the last term, since it is still relevant, and shows a part of the research this project will require.
And through the following one a miro board link collecting much of my visual research and the inspiration as well as quality check underlining this project.
I started with researching how things are done, looking at breakdowns, workflow and pipeline discussions, to figure out what I am planning to do.
Breakdowns
Breakdowns let me understand the layers of the simulation and what I require for each step, as well as the way the sim works and the differences between approaches depending on the project in works.
After some initial research I’ve learned that muscle sim does not limit to one approach only, there is at least two main ones that are still used (ways of making the sim not tools, there is loads of that). The approach we see most often, and the more impressive one, out of the two, is the full anatomy, multiple layer simulation (multiple muscles at one time working independently, with the addition of fascia, fat and skin layers), but there also exists single layer muscle simulation. This is based around using a single layer of mesh, then simulated to imitate the tissue secondary action (a bit like a softbody sim). It is used mainly for creatures or characters with no defined muscles structure, furry ones, where fur covers all or most of the detail or/and ones with a large layer of fat that covers all possible muscle detail, like in a polar bear.
An example in the fourth picture (link) in the breakdown section (above one), both the bear and the squirrel use single layer muscle sim to add simple secondary action where a full anatomy approach with muscles clearly defined would not be correct. In the case of a polar bear defined muscles simply look wrong, because thats not what we know from real life.
This approach would be cheaper, but it looses a great deal of control over the simulation, as the parameters work on the whole mesh, and despite the ability to paint them only in certain areas, we are working with one set of them and on one mesh, as opposed to multiple ones.
It might be good to explore that approach too if I have enough time, but my focus is on the full anatomy sim approach, at least for this project.
General, pipeline, workflow, improvement and experimentation
General, workflow and pipeline discussions gave me some insight and basic understanding of the process and possible issues and limitations.
What’s New – release notes & overview
Launch webinars and workshops often explain whats new, why and how to use it in the most basic terms, thats why I find them very useful and insightful, as in short time (unless specific like ziva ones) they simply explain tool mechanics and purpose.
Below, some ziva and houdini 19/19.5 launch webinars.
Tutorials
A. Ziva Dynamics
Manual
Single layer simulation tutorial
Multi layer simulation tutorial
Below picture shows the proposed workflow of Ziva use.
B. Houdini 19/19.5
Last term, when I couldn’t get Ziva to work, I tried a houdini workflow (that I also had issues with, and I know what caused them, don’t know how to fix it though), Originally I planned on trying it again while trying to figure it out, but then I found out that houdini is deleting in in the coming versions, as they are working on a new, more user friendly and universal approach (not object level rig dependant full anatomy process), so I will be trying this one out, if I have the chance.
The new system is around for only two versions and is still in beta, which means limited tutorials and documentation, but I actually managed to find some sources that basically teach everything step by step.
My issue however lies somewhere else, I tried to test it out and try to figure out the workflow recently, but after setting the sim rig up and playing the sim, the calculating took very long time. I tried to debug it and I figured out what causes it, but there is nothing I can do about it. In essence, I think my computer doesnt support the workflow requirements and if I cant get access to houdini in another workstation, this will not be possible as of now.
I will still list the research I have manage until now, but most probably from now on I will concentrate on Ziva and hope I can troubleshoot anything that I encounter during the project.
Other programs (in case and by the way)
During my research I have also learned that muscle simulation does not need to be done in a particular software, made for that purpose, of course programs designed for it will make it easier and will collect all necessary tools within them, but the task can also be done in basic 3D software of choice. I came across a few tutorials in a variety of programs, and I will also list them here, in case my program of choice doesnt work or if I have additional time after finishing the project in which I can explore other possibilities.
Blender
Maya (no plug-in)
3DS Max
NCloth – (membrane)
Another thing I will briefly have to research (I know the basics already) is cloth simulation. If I end up working with the idea I have chosen, I will need to figure out how to rig a pair of bat wings, and from my previous work with NCloth and Vellum Cloth, I am guessing that it will be done through the use of these tools.