View Full Version : rigging low poly to control a high poly?

08 August 2007, 11:21 PM
I know I have seen it done before because I've messed with a couple rigs that have implimented this feature, but what I want to do it make a character that allows for a easy and quick workflow.

The character I received has too much geometry and and would run too slow to interactively animate.

I have a low poly version of the model and I would like to find a way to rig that model and have the high poly on a seperate layer so I can animate the low poly and have the animation transfer to the high poly when I make the layer visable again.

Any suggestions?

08 August 2007, 03:45 PM
Two options:

A. The Uncle Ben's Low Res Rig (fast and easy)

Parent constrain the low resolution rig pieces to the appropriate parts of the high resolution rig. Store the low resolution geometry in a group and put that group in the low resolution layer, then stuff the high resolution geometry onto its own layer.

B. The Gourmet Chef Rig (difficult but production friendly).

Create message attributes on all your joints, define a 'to' and a 'from' attribute for the message attrs. Save your high res rig and your low res rig as two independent files and generate a script that looks for the message attributes when both rigs are present in the scene and runs a series of point and orient constraints on the joints using the 'to' and 'from' message attributes to get the appropriate attachment points.

Go with A unless you've got the patience of a Shaolin Monk and plenty of Rogaine to regrow all the hair you'll be ripping out :)


08 August 2007, 02:01 AM
You could simply skin the two models to the same skeleton and assign each mesh into a seperate layer (animation vs rendering) and then just hide the high resolution layer while animating and the reverse for rendering.

08 August 2007, 05:01 AM
well, production friendly in an environment where TDs are running the show. thank god they threw in trax and the hair system and ncloth for the rest of us..

i have found that using regular maya tools wisely you can acheive many of the benefits of approach B, while not doing all that much more work than approach A. basically you just need to design a consistent rig system for your whole project. or i use the same rig on pretty much every project.. and make sure its designed to work well with trax.

thats for another discussion, as none of this in particular has to do with changing resolution of characters. but i just had to mention it since ya went there. i've had stuff like that cause huge headaches on productions. keep it as simple as you can unless you are a real genius, otherwise the rest of the team is gonna hate you :) lengthy rigging nightmares that eat the whole schedule make animators have to crunch even longer and rush the results.. and animators love to polish..

08 August 2007, 04:26 PM
Well, B's the most solid method for a typical film pipeline to alleviate the artists from having to deal with large loads of simulations or heavy deformations. For our animators here it's a blessing as I'm able to simply chuck them a blocking rig to do all their animations with and the lighting department can work parallel to animation while the setup department finalizes the high res deformable mesh.

The less technical garble I can keep from an animator, the better the production runs as I'd rather have them setting keys than having downtime to watch youtube.

08 August 2007, 05:11 AM
i pretty much freelance on TV commercials or a shot or 2 of a bigger project, solo to maybe 5 person 3d teams. and i've encountered people who try to implement a whole friggin film pipeline on a project with something like 3 artists, 4 characters and 2 minutes of animation. or a guy who spent a week rigging a character for one shot about 5 seconds long, adding all sorts of cotrolls that would never be used.. we are doing one shot here..

the LESS paralellization of your process you can get away with, the easier its going to be to implement the pipeline, as far as i can tell. model, rig, animate sequentially is much easier process to plan and implement than paralell and requires less overhead (managers or dedicated TDs often un necessary) so thats usually how it goes (more or less.. client revisions, argh) on the gigs i work on. however its in this situation however that having a crazy TD that over rigs really slows the whole show. obviously on bigger projects you need to go in paralell and complexity shoots up and thats why movies take 300 people.. but as they say when you have a hammer everything looks like a nail, and i have worked with a couple too many TDs that seemed to just enjoy complex rigging and making pipelines than nobody but them understands, for its own sake, regardless of if its necessary..

i'm not saying really awesome and complex pipelines dont have a place in real high end production but just cause feature movies do it that way doesnt mean its the best in many other situations.. you have to consider schedule, scope of the project, staff, etc..

08 August 2007, 07:46 PM
well for a 5 person team or small projects it's pretty difficult to run a parallel pipeline given the fact that you're not necessarily working with a large pool of resources so yeah trying to jam a film pipeline into a small television production is like trying to jam a wrecking ball down a garden hose.

parallelization works on large scale projects much better than on smaller scale gigs simply due to the amount of resources and time differences. For me, I can't afford to have a team of animators on 'down time' during a project when I could be having them do their blocking animation using the initial silhouette model that the art director approved. Parallel allows things to run syncronously and also forces artists to pay more attention to changes they would make that could potentially cause a domino effect downstream.

There's lots of people who will overcomplicate things for the sake of bragging rights or getting to stake claims on the next technological advancement but a pipeline is only as good and efficient as the artists who use it. Like you said, if half your team can't wrap their head around the process then the pipeline has failed to do its job in making the transition from departments as seamless and pain free as possible.

I prefer to make a pipeline that we can bring a student in who's fresh out of school, sit them at a workstation, introduce them to the toolset/workflow and they'll be off and running. "Keep It Simple Stupid" is the name of the game when it comes to being a TD and anyone who ignores that general rule of thumb often suffers some very nasty consequences from surrounding artists :)

08 August 2007, 09:57 PM
Either use wrap deformer (works pretty well if the low poly is good) or just use substitute geometry when its final, and then tweak the skinweights some extra when you've transfared the high poly to the low one. :thumbsup:


08 August 2007, 11:18 PM
OK, I've been doing this for a fair number of years... and every now and then someone says, 'just wrap deform it'! I have never gotten a wrap deformer to work on a complicated mesh with satisfactory results. I've gone through the all settings I can think of going throuhg and dug as deep as I can but still bleah. Could someone fill me in on how to get this thing working well? Or is it like I think, not a great tool for precision work?

There's lots of people who will overcomplicate things for the sake of bragging rights or getting to stake claims on the next technological advancement but a pipeline is only as good and efficient as the artists who use it.

You are soo right on both accounts. I like simplicity. I'm a simple kind of guy with a simple kind of guy mind. Most of the stuff that people over complicate to the nth degree can be handled so much more elegantly.

08 August 2007, 01:00 AM
back when nurbs patches characters were still being made... it was great for that.. other than that, not sure..

CGTalk Moderation
08 August 2007, 01:00 AM
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.