Questionnaire on, What’s the most efficient method within facial rigging?
I'm a student of City Varsity, I'm currently doing research for my thesis on facial rigging.
It would be much appreciated if u could spare the time in filling these basic multiple choice questionnaire as it would very helpful for my survey.
Link : https://www.surveymonkey.com/s/WBXVK9L
It would be much appreciated if u could spare the time in filling these basic multiple choice questionnaire as it would very helpful for my survey.
Link : https://www.surveymonkey.com/s/WBXVK9L
Comments
In particular, your survey's completely focused on tech, which is suggestive of what you believe is important to know/learn. Judd advocates having strong art foundations, because of how a very basic rig that is able to create appealing deformations is infinitely more valuable than a complex rig that simply looks ugly.
Good luck.
Also, the problem with the question you are asking as that efficient is quite a loaded term. What is efficient (how quickly you can get something done), may not be effective (getting the job done well) for a given situation. A rigger will always need to balance what is efficient vs what is effective
Apart from that little side track, usually dependent on the task at hand you will combine several techniques. For example joint based facial rigs, with corrective blendshapes on top as per most modern AAA games see: http://www.chrisevans3d.com/pub_blog/multi-resolution-facial-rigging/
for a more in-depth example.
Facial Rigging can be broken down into a few areas, games/ film and advertising, and even then in the scope of film and advertising there can be a great difference in what is needed and what you want to achieve.
What you have asked can be compared to asking what is the most efficient way to build a building, and not how do you build a skyscraper, or a garden shed, there are two many unknowns in your question.
Most AAA games are a mixture of bones and blendshapes as shanemarks and Elyardine have mentioned, but that also goes to any facial rig setup.
What area are you wanting to make the facial rig for? mobile, AAA, film?
What is the turn around time for a face rig? are you one person, and you have a 3 months to just make the blendshapes, or do you have 1 week for an entire facial rig.
What system / pipeline/ game eninge are you using? can it support that many bones, shapes and so on. Unity would die a horrible death with that many bones and shapes in the example that shane posted. A film studio would most likely send that facial rig back to you, as they would have active deformers and going on 500+ blendshapes and correctives depending on the studio. as each has a different pipeline, and depending on the crew size and availability of time.
I have worked at some studios where you have a week for an entire character, including face, and others where you have a few weeks per a character, and I know of other studios where you will only do the blendshapes and you have about a month on one character.
With all my rambling a facial setup comes down to what do you want to achieve with the facial rig, and does it achieve what you want, and look good doing so.
You should try and focus your questions a bit more with regards what kind of data you want to get out. Asking a rigger which system they use for making blendshapes, is rather superfluous as it doesn't really matter. In the end the studio you work at will depict the software you use, as they will have licenses, and faceshapes are just a weighted vector which push's out a normal. We have written small applications to created blendshapes and then you can export the weights out and into any other application.
Questions 4:
Baked Animation on the Mesh (Point Caching) is only ever done in film for rendering, as you wouldn't stream point positions onto multiple mesh's in a game, the overhead would be huge.
You left out the most used method of animation importing into a game engine, FCurve Baking - allowing the basic underlying rig to just be imported into the game engine, and not bringing in all the ik/fk/hierarchy data, this allows for lighter scenes and faster evaluation, but also this is done more for mobile games, and generic cycles.
Question 5:
What are you referring to, its a very open ended question, there a tons of new theorems with regards to new binding/enveloping techniques and methods, there is also alot of new Operator API's for faster rigging, but this all relates to the rig creation/animation stage and wouldn't transfer into a game engine, as the game engine would have its own set of operator evaluations.
Question 7: Comes down to the application used really. The mass standard for scripting in most studios is Python, and in Maya python is basically just a sh*ty MEL wrapper, its not very object oriented. Also you wouldn't code in MEL in softimage/blender/lightwave/3dMax as it is only available in Maya. Rigging being a technical job also relates very closely to the application you are rigging in, you would rig differently in Maya then in XSI, as the applications have different strengths on weakness's, its not like modeling or texturing.
Question 8: what has this got to do with Facial rigging thesis? The reason I ask this is most basic knowledge and understanding comes from tutorial sites, and then the rest comes from blogs/user videos, Siggraph papers and other research outlets, and messing around with techniques you have seen, and trying to recreate them.
My suggestion to see and find out more about facial rigging is go trawl rigging blogs and vimeo rigging account, there is usually alot of information, and break down of setups, that can help in putting the pieces together.