Poly Counting

 (For all images below, feel free to click on them to view the full sizes)


I. Poly Counts

This Past week was spent with a lot of time looking at ways to reduce polygon counts of the fossil scans in order to make them useable for animation without crashing every program known to man. The full resolution scans are priceless for the scientific community, but are very impractical in an animation production environment. I was worried I would have to block model every bone to create low resolution versions to use, but luckily, there is a tool inside the scanning software (Geomagic Studio) that allows a user to specify the number of polygons to reduce the geometry to. Based on past characters I have built, I estimated that around 10,000 triangular faces is a good number to use to keep a lot of the detail while reducing the number of faces significantly. The original scans were coming in around 400,000 tris...so reducing to 10,000 is an extremely significant reduction (About 1% of the poly count of the original) but doesn't take away too much of the detail.

  

 

II. SKELETAL RECONSTRUCTION

After I created a series of lower resolution versions of each bone, (10K tris, 5K tris, and 2K tris) I started to put the skeleton together. I used the procedural nature of Side Effects Software's Houdini in order to allow me to quickly position the low resolution versions, and create a switch that automatically swaps out the geometry for higher resolution versions. I will soon be looking into creating some sort of script that allows me to swap the 10K resolution bones to the full resolution bones at render time, so that the computer doesn't ever have to display them in the view port, which crashes the program every time.

 

...and just for kicks, here are some other views of the 10K resolution:

III MISC.

 I've also done some more R&D for muscle systems and for environment development....more on that to come soon. I have gone through the whole pipeline for Maya muscles a couple times to familiarize myself with that, and I have been putting myself through a crash course in Real Flow for water dynamics. Stay tuned for updates on that.

On a side note, in my continued research for croc anatomy, I came across the relatively recent British television series, titled Inside Nature's Giants, where a team of scientists dissect a different large/specialized animal in each episode. One episode happened to be about the Nile Crocodile! This was very informative in not only showing the musculature and other structures underneath, but also putting the anatomy in a physiological and evolutionary context. It was extremely effective and I learned a lot very quickly. I strongly recommend it to anyone who's interested in this stuff. The other episodes were pretty great too (I especially liked the one about the giraffe).

Knee Deep in Literature

So lately I've been delving deeply into the Literature, trying to resolve some of the issues with the missing skeleton parts.  To get my feet wet, I read through the skeletal portion of Robert Chiasson's Laboratory Anatomy of the Alligator. After that it was an in depth looks at some Christopher Brochu Papers, including some about crocodylian phylogeny issues, and some papers about Eothoracosaurus and Eosuchus. Unfortunately I was having trouble finding any papers dedicated to Thoracosaurus neocesariensis in particular. Every lead I had pointed back to a Kenneth Carpenter paper on the species, but I have been having trouble locating a copy. That being said, Brochu describes the same specimen that Carpenter does, and argues that it is in fact not actually Thoracosaurus, and distinguishes it as Eothoracosaurus. In these comparisions I was able to find out some descriptions of Thoracosaurus, which was very helpful.

For what I cannot find on Thoracosaurs neocesariensis I will have to extrapolate from the other "thoracosaurs" (Eothoracosaurus, Thoracosaurus, Thecochompsoides) as well as the modern gavialis.

I then did some reading up on Gharial descriptions and behavior. I found some great information on their hunting behavior in particular, which was very interesting. I am very close to being able to draft up a proposed animatic based on the behavior of these large fish-eating crocodylians. I have also been looking for video documentaries on gharials to watch and study, so I can see some of this described behavior in action. BBC motion gallery is a great reference but only goes so far.

I am now working on a large variety of things. I need to jump head first into the muscle systems. But more importantly I need to figure out the best way to create low resolution "proxy" bones to act as stand-ins for the high resolution scans, so that I can actually rig and animate them. That's the next thing on my list. I may have to hand model some simple proxies, and the scanning software (Geomagic Studio) may be able to do a lot of the heavy lifting for me. Then I'll have to write some sort of script that substitutes the proxies with the full quality at render time... Lots to do lots to do! Stay tuned for more updates. Once I figure out this pipeline, I promise I'll post a rendered image of the full digitized skeleton.

Literature Read this Week:

Brochu, C. A. (2006) Osteology and phylogenetic significance of Eosuchus minor (Marsh, 1870) new combination, a longirostrine crocodylian from the Late Paleocene of North America. Journal of Paleontology 80(1):162-186.

Brochu, C.A. 2004. A new Late Cretaceous gavialoid crocodylian fromeastern North America and the phylogenetic relationships of Thoracosaurs.

Brochu, C. A. 2001. Crocodylian snouts in space and time: phylogenetic approaches toward adaptative radiation. American Zoologist 41:564–585. 

A study of fossil vertebrate types in the Academy of Natural Sciences of Philadelphia: taxonomic, systematic, and historical perspectives Issue 16 of Special Publication Series, Academy of Natural Sciences (Philadelphia, Pa.) By Earle E. Spamer, Edward Daeschler, L. Gay Vostreys-Shapiro. Academy of Natural Sciences, 1995

Thorbjarnarson, John B. 1990. Notes on the Feeding Behavior of the Gharial (Gavialis gangeticus) under Semi-Natural Conditions. Journal of Herpetology 24: 99-100.
 
Whitaker, Romulus. 2007. The Gharial: Going Extinct Again. Iguana 14: 24-33

* The list on the right hand side is what is next in que for the research aspect of the project. If anyone has any suggestions of papers to read, please let me know and I'll incorporate it into my que!

Oh yes, and I appologize for the lack of pictures.

 

Look What I Found!!!

Hey hey hey, I was able to dig this bad boy out from the depths of the scanning laptop.

Putting it together went really quick too...I think it's because A.) I am getting much faster at this thing....and B.) There are lots of pointy bits on it that make it really easy to match. Very exciting...because this means that SCAN STITCHING IS COMPLETE!!! Now....as soon as I can figure out how to get all of this geometry to render without crashing any of the 3D packages, I'll be sure to render an image of the complete fossil material. In the meantime it means I need to do a lot of reading so I can start modeling the missing pieces...oh yes, and muscle system research has been going on as well.

Closing in On the End of Scan Stitching

So I'm very close to posting an image of the fully digitized fossil material. There are a few pieces that a colleague of mine is finishing up, and I need to locate where the left dentary has run off to... It appears to not be among the files I've been working off of. There's one other place it could be, which is on the laptop that was used when scanning. I hope it's there so I don't have to rescan anything. There is also one dorsal vertebrae that seems to be missing data for about half of the bone....I am going to look into stitching an older scan of that bone together, from when we were using PET opposed to Geomagic. It will take a lot longer to stitch, but It would be better than having to schedule an entirely new scanning session.

In the meantime, here is an image of the skull. The skull was incredibly difficult to stitch together, since each fragment was very difficult to scan. Some of these pieces are being held together by a very thin layer of Paleobond glue. Because of the fragmented quality of the skull, as well as the complex shape in general, it was very difficult to get a complete scan. I worked with what we had and was able to get them pieced together as best I could. The skull is so fragmented though, and a lot of the detail that the scanner didn't pick up consists of the interior of breaks...i.e. material that wouldn't be visible if it was a complete skull. When the full skeleton is built, and the entire skull reconstructed,  it will not even be noticeable that some of these crevices were not picked up.

In the meantime, it's on to digging through the literature so I can start modeling the missing bones.

Large Alligator Motion Reference!

Great day today! Jason Poole and I drove up to Allenwood, PA to visit Clyde Peeling's Reptiland. There were lots of cool things there, and they showed us around a bit behind the scenes of the place (including off-display animals, and the shop where they build all the custom habitats). It's a really cool place and I highly recommend it to anybody interested in zoology. The best part of the day is when we were able to shoot some video footage of their 11.5 foot alligator, Rocky. We had two cameras and were able to capture the large fellow standing up, turning around, and slumping into the water from two different angles. The weight of this guy was very impressive. This should prove to be very helpful for the Thoracosaurus motions, since they are both very large crocodylians (the Thoracosaurus is estimated around 14-17ft long). I also shot some additional footage of the gator's skin for texture reference. The staff was all very friendly and very helpful. Thanks Clyde, Chad, and staff for letting this happen!


Alligator Motion Reference from Evan Boucher on Vimeo.


Also, Many more stitched scans to come later this week...

More Scan Stitching

I just spent a lot of time getting together various bones. Below is a sample image (click to enlarge). I put together the humerus, femur, a couple "scutes," a tooth and started on what we have of the skull. The skull is in five different pieces. It is currently cleaned up but just needs to be put together. After the skull and jaw are together, it's nothing but vertebrae left for the scan stitching.

I was able to vastly improve my workflow with the help of a "Global Registration" option in Geomagic, which allows me not only to speed up the process of stitching, but also helps create more accurate piecing, since it is not done by the human eye. As long as the pieces are in close approximate positions (done manually), the algorithm is able to detect the topology and adjust itself for a more perfect fit.

Despite this discovery, it took a little bit of playing and experimenting to figure out how it works. There were lots of little strange bugs that arose when using the program that slowed down my newfound speed, but each day I am getting better at this.

I've also started reading my que of papers on Crocodylians by Christopher Brochu. This week is: "Crocodylian Snouts in Space and Time: Phylogenetic Approaches Toward Adaptive Radiation" (2001). This will be helpful when reconstructing the missing snout of our Thoracosaurus.

I also spent time on looking up reference imagery of marine forests, mangrove swamps, etc. to influence the revision of my environmental plan. Stay tuned for some concept art on that, as well as some beginnings of possible storyboards of the animation.

Proxy Environment Startup

I blocked out an environment to help me start to visualize the final result for when I storyboard... Here's the super simple mangrove swamp, complete with trees, bushes, and even a proxy thoracosaurus...