Tyrannosaurus rex skeleton

[dinocarl]

Hey everybody,

 

It's time once again to post my latest work in progress, a T rex skeleton. For anyone who knows me (or even my handle), this won't come as the slightest bit of surprise. Several years ago, I completed a T rex skull, and I always wanted to complete the skeleton, especially since Joe Cosman's Albertasaur is such an amazing model and has long been a wonderful bit of inspiration. So here we go:

 

The goal at this point is get as many pieces made and in place. I am going for fidelity to my source material, but not extreme levels of detail. This is obviously a subjective standard, so I apologize in advance for inconsistencies along the way. However, because of the nature of the model itself, I can and almost certainly will go back at various points along the way to modify, add or even rebuild the various pieces that are already here.

 

I've already completed the arm and shoulder blade, six of seven parts of the pelvis, the femur, tibia and fibula. I've begun on the metatarsals, but it hasn't made it into the render, yet. I've actually already modeled a few of the vertebrae, but since there are so many (80, by my count) and each is a variation on its neighbor, they're going to require a special effort all their own.

 

The model is currently a little under 12,000 patches, with just over a third of that being devoted to the skull. This image took about 15 minutes to render at 640x480, with one light, radiosity and 5 passes.

 

Here's the render:

[robcat2075]

Fearsome!

 

I bet AO or FakeAO would do as well and faster than radiosity.

[jason1025]

cool man

[Rodney]

Sweet!

 

He is already looking great.

 

I've already completed the arm and shoulder blade, six of seven parts of the pelvis, the femur, tibia and fibula. I've begun on the metatarsals, but it hasn't made it into the render, yet. I've actually already modeled a few of the vertebrae, but since there are so many (80, by my count) and each is a variation on its neighbor, they're going to require a special effort all their own.

 

I'm curious and assume you have some great sources to model these from.

Just thinking about how the dinosaurs might have moved around makes has me interested.

 

The model is currently a little under 12,000 patches, with just over a third of that being devoted to the skull.

 

With the detail you've got there that seems very economical.

When this guy is done I'd sure love to see this guy added into a couple of Stian's highly detailed scenes... ahem... for T-Rex to modern day structures and conveyance size comparison. Hehe!

 

Note: In writing this response you've just made me realize that I have never once considered a T-Rex as a female... so you are already educating me!

[Rodney]

If you haven't visited Carl's website (or haven't been there in awhile) you really should drop in and visit the site:

 

Here's a cool example of Carl's animation:

http://www.dinocarl.com/animation/homage-to-jurrasic-park

[Vertexspline]

Nice Model Carl....and your right Rodney --Carl's website is worth a look.

[NancyGormezano]

Wonderful! Terrific artwork at your site.

[robcat2075]

I've actually already modeled a few of the vertebrae, but since there are so many (80, by my count) and each is a variation on its neighbor, they're going to require a special effort all their own.

 

Thinking out loud here...

 

I bet there are a few vertebra that are like "keyframes" with pretty much straight morphing inbetween them.

You could make one master vertebra, with a pose slider to transition it through the key forms.

 

Use Multiple Models On Path to distribute a bunch of copies along a spine path then step through them to tweak the pose setting on each copy of the model.

 

Quick example...

 

[dinocarl]

Thanks for the positive feedback about the model, and for the kind words about my site, Rodney and Nancy. Rodney, one of the sources that I've been using is the Black Hills Institute site - they have a number of amazing replicas for sale, including a full T rex (for the modest price of $100K). They sell 2 different arms, one gracile and one robust. The gracile they attribute to a male and the robust to a female - the justification for which is that in most other animals, the females are larger. Of course, the most famous female rex is Sue, another amazing resource to have just a few miles down the road from me.

 

robcat, I considered something vaguely similar inasmuch as I had played with duplicator, but I think your idea is better. There are definitely the big landmark vertebrae that can be modified in the way you suggest. My only hesitation would be that I've found that overloading a path constraint can cause serious performance hits during animation, but I think the happy medium might be "modeling" the vertebrae according to your scheme, then exporting the action or chor as a model. Either way, thanks for the great idea and I'll make sure to post about it when I get there!

 

Carl

[robcat2075]

My only hesitation would be that I've found that overloading a path constraint can cause serious performance hits during animation, but I think the happy medium might be "modeling" the vertebrae according to your scheme, then exporting the action or chor as a model.

 

Yes, I only recommend this to make the model, not for rigging it.

[dinocarl]

Sorry for the delay in getting out an update. Holidays, job-hunting, head colds all played a part in prohibiting further work. For this installment, all I've done is add the top metatarsals. It's not much, and I'm terribly excited about moving on to the toes, but these were definitely tricky to model. Here's the update:

 

BTW, robcat, I meant to say it in my last post - you're right, an AO render would reduce render time and look largely similar. However, I do prefer the radiosity render and, given that I have two machines in my render farm, it's not a problem for me to send something off to them while I keep working on something else; it's not my fifteen minutes being taken up.

 

Also, because today's update is so meager, I thought I'd show off some of the rigging that I did for the model. Here's the arm moving about; it's nothing special, just a range of motion test.

trexarm.mov

 

Carl

[dinocarl]

For this update, I've added the 3 main toes; hopefully I can get to the 2 vestigial toes this weekend.

 

The model now stands at just under 16,000 patches. BTW Rodney, I am trying to be economical while remaining true to the overall shapes. Of course, the problem with almost every bone is that the great number of protrusions, dents and holes in/on them really get the patch ticker going. And I just did a rough estimate of what the vertebrae are going to entail, and they will likely run me another 30,000 all by themselves.

 

I haven't rigged the legs or feet yet, either, and that's high on my to do list; for this render, I just positioned everything manually. At any rate, it'll function as another procrastination strategy until I can no longer put off making those vertebrae.

 

robcat, after all that talk of not caring about render times, I've gone ahead and used AO for this render with 2 more lights thrown in for good measure. My radiosity set-up was washing out a lot of the detail (especially in the feet), so I switched, at least for this update. As predicted, my render time savings were about half.

 

 

Carl