I’m currently a Kinesiology major in university, studying anatomy, physiology, body kinematics and stuff like that. I’ve recently developed an interest in rigging and i’ve been messing around with some tutorials and stuff with the Maya educational edition. I was wondering how much that I have learned in school will apply to rigging.
I guess a better way to phrase the question is:
How much knowledge of anatomy is actually required for rigging?
Is there a specific specialization of rigging that requires more that I can apply my knowledge to?
From what i’ve heard Autodesk has stopped supporting (?) Maya Muscle which seemed pretty cool and applicable, but is there anything else that people actually use in the industry?
If the question is how MUCH knowledge of anatomy is necessary for rigging… the answer is actually very little.
A basic understanding of the overall musculoskeletal system, (and even then only the superficial muscles) is pretty much the limit of it.
The aim of character rigging in CG is simply to give the IMPRESSION of a living, breathing character (talking ofc in terms of photorealistic characters/creatures). Certainly knowing, and following the principles observed in reality is helpful, but ultimately… you go for what LOOKS good, and if that means violating the norms of anatomy (or even laws of physics)… then thats what you do.
“What looks good” is highly subjective.
Of course it’s possible for a rigger to bend a character around and be fully content that it’s doing what they want and is up their own standards and their clients.
However, What is generally recognised to be the high end creature work has been achieved by artists closely referencing and understanding the ‘engineering’ of human and animal anatomy and with teams RandD ing how to replicate all the phenomena that contribute. There is a constant push for more realistic internal materials/structures and interaction between them and how that affects the skin above.
Its a very different art to skinning with cgi bones and adding some morphs to fake the look at the surface.
The general Standard changes over time as the bar is raised and it gradually filters downs to the general cgi community. Right now we are in the creature equivalent of where rendering was 12 years ago. Ie… GI and more advanced shading etc generally not being used. but technically possible.
Well, it does help. Once you know how a shoulder works, how the radius and ulnar rotate your underarm and a ribcage bends and compresses…
…that IS helpful knowledge. Being able to apply anatomical knowledge instinctively and on point for rigging characters is a lifesaver. It helped me out in a lot of tight projects, where rigging had to be done yesterday and we had not a lot of time for testing.
Realistic muscle systems are slowly setting in, so your knowledge will become more and more valuable. However, in depth kowledge of the human anatomy will only be useful for certain areas, such as medical visualization.
On this one Im afraid Im gonna have to take a lil disagreement, lol… well, sorta.
Ofc you’re right that what looks good is subjective, but that was kinda my point. You could rig up a character with a FULL working internal anatomy, accurate in every last detail, and the result might be that it just doesnt fit the bill. it needs to be “bulgier”, or whatever. On the other hand you could have something that’s only a very loose replication of reality that looks amazing.
As for “what is generally recognised”… I personally think that’s something of an emperor’s new clothes phenomenon.
Go and watch some rigging behind the scenes where they show off just how intricate and accurate their internal mechanics were, and its pretty common that folks will THINK they’re seeing something better, or more realistic, when in fact, they’re not. It can easily be a false perception.
Honestly… Ive yet to see any demonstration, or paper or talk, whatever… that REALLY gets into the nitty gritty of it. Muscle systems are a classic example. The final shape we see on the skin is the product of the stretching and bulging of multiple layers of muscle… the pectoral and obdomen being great examples. Yet so far as Ive seen no-one bothers to simulate all this cross interaction (cos it’d be a nightmare no doubt, lol)… Most muscle systems rely on just shaping the superficial muscle layer and leave it at that.
Ofc, getting away from the sim stuff… there are other considerations regarding rigging which come down to the skeletal mechanics.
Lets consider the forearm/wrist as a classic example (and here Im gonna talk in terms of the “anatomical pose”, body upright, arms at sides, palms facing forwards)…
The pronation/supination of the forearm in reality occurs axially around the ulna… that is to say the ulna stays “fixed” in space while the radius curves over and around it. Assuming the arm is “locked” this has the effect of shifting the hand side to side a lil across the lateral/medial axis of the ulna. But this is almost never taken into account by most riggers, the most common practice by far is to effect the rotation centrally (as though around a rod between the ulna and radius).
this also follows into the wrist, as we tned to plonk the wrist joint right on the end of the forearm (one common reason for keeping the forearm central). But uh-oh… there’s a problem with the wrist too. Whilst extension/flexion of the wrist happens (more or less) around a central point in the wrist (viewed in saggital aspect)… the adduction and abduction of the wrist (side to side turn of the hand) does not, and the centre of rotation shifts considerably.
As a series of pivot points to setup (and slide around) this is a bit of a pain to rig, and again… ive yet to see many examples where anyone has bothered, other than as exercise/experiment.
Or there’s the knee joint. The tibia has rotation around a central pivot in its proximal head, but the entirety of the bone also slides around the distal end of the femur, which creates an eliptical rotation path… but again… this is almost always ignored, and in CG knees are pivoted around a single, fixed (relative to the femur) pivot point.
Then once u get through that there’s interractions between the skeleton and musculature… consider the bicep. As yu put flexion on the forearm, (with the palm still facing forwards), u get one kind of shape in the bicep ( a more “ball” like shape). But pronate the forearm, (with the elbow still flexed) and the bicep shape changes completely to be much more drawn out… This ofc being due to the 2 distal heads of the bicep attaching to the proximal ends of the ulna and radius (which are twisting around one another). You also, if u look carefully, will note that this “stretching” of the bicep occurs mor notably on the lateral side on the muscle than its medial side… again, due to the action of the radius.
But again… u never see or hear of this being taken much into account in muscle systems, and tbh… Im not surprised. Considering its one change effected by one axis of rotation on just one lil bit of the mesh, its gonna be far easier to just bodge that lil bit (should u even want to include it) with a morf or similar.
Plus ofc there’s the fact that anatomy in nature isnt really “fixed” so to speak. There’s a fair amount of variation between bone thickness and muscle mass, proportions shift around… hell even the number of vertebrae is variable… so when confronted with a given character mesh, what its interal anatomy SHOULD be is pretty open to interpretation (and artistic licence) already.
All in all… if it looks good, and looks real… no-one will care about the glitzy details of how it was made.
All in all… if it looks good, and looks real… no-one will care about the glitzy details of how it was made.
Studios dont just make anatomical systems for the hell of it. Its a lot of work. so again… it boils down to the subjectivity of what ‘looks good’ and how far one wants to go to achieve that level of realism.
All the time, the bar and general standard is raised higher. Look at the general standard of ‘photoreal’ humans from the nineties and compare that to today.
There is a huge shift in the acceptable quality of character work approaching. As processing enables more complex systems to be built. As creatures and especially digital humans are the final visual frontier. many studios are upping their game and this is filtering down to the smaller studios too, to keep pace.
Muscle systems aside, I would say that whatever you use, yes its very important, if you want to make something look accurate, to understand the real world phenomena that you are trying to fake. Of course one doesn’t go to the extreme of studying the inner workings of the pancreas and liver etc. but that is becasue they dont affect the outer shell. Its beneficial to know as much as you can about everything that does though.
Oh no, ofc… no disagreement there. these systems DO most certainly improve the eventual look at the skin surface. But my main point was that they dont need to be all that complex, or highly accurate, and really only need to take into account the most superficial anatomy. But I would have to say that making them more and more anatomically correct/accurate is a race toward diminishing returns, and that a kind of “delusion” (confirmation bias) develops where because someone has created something more intricate, they falsely believe it looks better.
If u were to take one character setup with a basic such (near as dammit) system, and another with a much more complex (or “accurate”) system… I personally would defy anyone (who wasnt themselves well schooled in such anatomical things) to tell the difference just by looking at the final skin effect.
So basically, you are saying that , assuming it was totally photoreal, shading wise someone couldnt tell between real life (which is the usually the goal of these systems ), and a rigged character, using morphs and bones?
In which case it would be an enormous waste of time to develop these systems, yet that is what many of the big character studios are doing.
I can see a huge gulf between the current level of CGI and real physical bodies and I’m sure nearly anyone could, given a side to side comparison.
No, I didnt mention “looking at real bodies” Im saying that the vast majority of folks couldnt tell the difference between a character skinned with a system that relied on a basic, “near enough” anatomy approach, and one that used a highly complex and accurate system.
Im saying that going after more and more complex systems to simulate anatomy provide diminishing returns. All you really need is the most obvious, superficial structures. Delts, biceps, abs, pecs, etc, and a couple bony protrusions with a bit of skin sliding.
I have to agree with that, the amount of work and effort put into those rigs doesn’t really translate to a 1:1 effort/quality. sometimes something much simpler with some clever tricks will look as goog or nearly as good, specially if you are the average guy who goes to watch a film in the cinema and don’t come from a background on cgi…
I agree that if you made an anatomical system from scratch for every character, that would be an outrageously unreasonable amount of effort. Luckily, that doesn’t have to be necessary.
Even though it seems like discourse, there is a rough agreement in that the effort should be put into the stuff that gets the most time on screen.
To me it’s like doing life drawing, you can draw a person that looks good by drawing what you see but if you have a good grounding in anatomy at a muscular level you can structure your figure better and make it “feel” more correct. Anatomical knowledge is definitely a requirement when you’re doing anatomical deformations but you wouldn’t need to know how the nervous system works though. It’s main strength in doing deformations is that it makes it easier to see when something doesn’t look right and why.
In rigging how you put that anatomical knowledge into practise is up to you. There are plenty of ways in which you can try to achieve the end result - full fusiform muscle systems, skinning with pose space deformers and/or just sculpting detail shot by shot. Bigger studios have R&D departments which will write their own anatomical deformation systems (but that doesn’t mean everything gets done for them!) while for individuals and those in smaller companies it’s often best to put the work into the stuff that actually gets seen in the end …because it’s usually a lot of work.
Data is definitely moving into the computational arena, rather than a manual one - for example solving facial blendshapes in a pose space, rather than hand-crafting and knowing how a mix needs to be properly interpolated. Solving issues computationally has two benefits:
firstly input data can be hand-crafted, and secondly solved data can be ‘tuned’ aesthetically.
It’s not really a case of using fusiform approaches over pose space deformations - its more a case you want to ‘solve’ the variety of unknowns that the character can do - unknown poses, motions etc that need to hold up on the screen or in the game. And the ability to ‘craft’ exact shapes for hero frames/shots.
Building a full muscle system is a huge endeavor, and what tends to happen is a combination of a broad muscle system with PSDs and shrink-wrap skin sim’s for fine creases.
A solved system is nice in that it can be very finite - i.e only solving for certain situations or very broad i.e every possible combination of a rig, slider bank etc.
Faceware by example is doing this - when the talent makes a smile pose in the video, you make a ‘key’ smile pose using sliders and faceware then solves for the whole video. The benefit is again tune-ability and the fact that after the data’s been solved you can still hand key poses and make refinements.
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
