Jul 4, 2009

Already have my nose to the grind shale

Well Septemeber's ART Evolved gallery is centered one of my all time favourite critters, the Anomalocarids. They're not exactly a household name, but I think on many levels they are just as cool as T-Rex (which to put that in context, my favourite Dinosaur is T-Rex!).

I'm planning on going all out for this gallery, and am aiming for at least 3 new pieces! Not small scale ones like my Pterosaur one either (which was pretty pedestrian, apart from making the 3D wings flapable) but more in line with my flukly awesome Gorgonopsid one.

So the first two tasks I set for myself this week are to improve both my existing Anomalocaris model and my water effects. After that I move on to modelling the environment and other prey critters.

Without ego I will state my water effects are one of the strongest things I do, and are worthy of a low budget Finding Nemo (Nemo being the greatest 3D inspiration I've ever had, but yet my greatest nightmare! It hurts my head to watch it with a 3D CG sensitivity. Yet Nemo's "making of" extras have given me so many ideas and hints!...).

In this test render I've made a minor alteration of my water "sunlight" effect. The improvement aren't as obvious in this render, but once I add the Cambrian "forest" underneath you'll see how cool the rippled light shafts look.
My first real progress has been on my Anomalocaris model. This pic here is of my original 2003 version. It received a few modifications in 2007, but they were just colourization and the two long tails at the back.

Considering how much my 3Ding has improved in the last couple years, this is one of my strongest old skool models. I haven't had to change about 50% of it. Though in fairness Anomalocarises are nice in that they have fairly simple body forms.

The two biggest problems with this guy were the "arm" tentacles at the front of the head and the body.

My tentacles though conveying the anatomy, are utter rubbish. This was mostly due tolaziness of younger me. I didn't want to build the tentacles out of multiple segments as back then they'd take my much longer to pose. I didn't have a grasp on how to link and group models for proper posing in those days. I would have to move each and every piece as they won't move with other pieces they should have been connected to... which isn't easy to do in 3D as you might think!).

So this models tentacles are simply a single tube like object I've modelled into a curved shape. To make life easier (again I was trying to be lazy) I tried to invoke the illusion of segments through a shader with lines. Then to get the trademark spines on the tentacles I added them via a surface replicator, which though made them quick and easy to reposition if I changed the tentacles position, made them rather unrealistic.

The body though somewhat functional as is it here, no longer matches what I think these animals should look like. In typical restorations following the trend set by Marianne Collins' landmark Laggania restoration these animals have been shown to have large almost fatty bodies. Which this body matches (other then lacking a fating of the body in the middle as seen in the fossils).

This to me makes little sense as specimens show they clearly had flexibility in their bodies, but yet having an exoskeleton this shouldn't be possibly without built in segments or joints. I intend on doing a big post about this topic on ART Evolved in a week or two.

So here in the new version. I have completely rebuilt the body so it is a sequence of 14 separate segments, each corresponding to and housing a pair of the fins. The tentacles I'm most proud of. Not only are they made of the proper 13 segments, but each now wields the correct pair of three pronged spines, and the whole thing is a unified linked model.

With my new mastery of of groupings and rigging this guy if fully posable without the need to move and position each individual piece. Rather he is a link form model. This is different from the skeletal rigging I've been ranting and raving about here lately.

Linkform models are put together as though they are a machine or action figure. Each individual part links to the next with a natural rotational point created inbetween them. This is different from a skeletal rig which puts bending joints into the middle of a single 3D object.

Though I could probably mimic the effect of an exoskeleton with a skeletal rig, this seems to me a LOT more work and precise setup, where the modelling of an exoskeleton lends well to simply using links to achieve the perfect effect. I'm preparing a series of posts on ART Evolved about various means of posing and rigging 3D models (due to the misadventures setting up my Pterosaur) if you don't follow what I'm saying here.

1 comment:

Sean Craven said...

So is the tail a default for anomalocarids? I mean, is it most conservative to assume there is a tail?

Because I ain't got one on my A. canadiensis stuff...

And those feeding arms are a bear. They're the most complicated part of the body and all that toothy detail is packed into such a small area. I'm having to do separate enlarged sketches and then composite them back into the original sketch.