On VR, Scale Visualization, and Really Big Dinosaurs


via markwitton-com.blogspot.co.uk/2014/09/hey-dreadnoughtus-not-so-close.html

At the behest of my advising faculty, this blog has been reincarnated as a hub for my thesis research and production updates. Don’t mistake my summer hiatus for a lapse in productivity — I’ve been crunching away at my paleoart reconstruction work, but I’m still awaiting information about what I’m allowed to share publicly. Rest assured, cool things are still in the works!

A brief recap: For my Master’s thesis, I’ve been working on an animated reconstruction of the titanosaur Dreadnoughtus schrani, one of the largest creatures to ever walk the earth. This project began as a digital paleoart animation, but has evolved to suit a medium that is rapidly gaining traction in contemporary digital media: Virtual Reality.

While researching scale within the the context of conventional cinematography — achieved through camera angles, frame composition, careful lens choices and selective editing — it occurred to me that this language is absent in the emerging medium of spherical cinema. While VR device ownership is on the rise, interactive spherical videos are becoming more accessible than ever. Facebook now allows users to share 360-degree videos, which use a smartphone’s accelerometer to control the POV within a dome-shaped video scene. 360-degree YouTube videos can also be viewed through the lens of Mobile VR devices like Google Cardboard, conveying an unprecedented level of accessible immersion to an audience of social media users.

When the viewer controls the camera, the director loses control over scale relationships on a shot-by-shot basis. This not only affects how big something LOOKS, but how big something FEELS — an important distinction in any narrative medium. Preserving accurate scale representation is specifically important for Paleoart, which strives to portray extinct animals with the highest degree of accuracy possible (within the limits of scientific literature). When your subject weighs upward of 60 tons, scale visualization becomes an essential factor in this portrayal. Dreadnoughtus could tip the scales against an entire herd of elephants. Most people have never even seen an elephant in person — how can you picture such an animal in your mind’s eye? Creating an educational Dreadnoughtus film for Virtual Reality could facilitate the public’s understanding of the animal, as evidenced in similar studies that suggest that immersive media technology can influence an individual’s ability to remember information in the physical world.  

When communicating scale in virtual reality, film language is substituted with the sense of “Presence” induced through spatially immersive environments that blend the boundaries between “real” and “digital.” This evokes the same visceral feeling one gets when looking up at a mounted skeleton in a museum, and can be achieved through careful environment design and narrative cues. My thesis now takes place over three immersive environments with differently created background plates: Photographic, Videographic, and Computer-Generated. Whether the complexity of the environment has any effect on sauropod scale visualization is yet to be determined. (Storyboards […storycircles?] to come in the near future!)

Presence in realtime media implies a “feedback loop” between the player and the experience, which makes the player feel like they affect their surroundings. The limitations of cinematic VR restrict interactivity, but generally allow for higher graphics quality to be rendered. For some purposes (architectural visualization, for example), real-time rendering engines can achieve photorealism that rivals rendered CG. It’s easier to render “things” than “creatures,” though, and the Dreadnoughtus rig will include softbody muscle simulations that are much harder to achieve with a realtime engine. Rendering this project as an immersive video also dramatically extends its reach, allowing it to be shared via Facebook, YouTube, and Mobile VR, and (potentially) educational programs like Google Expeditions.

I’ll be updating this blog weekly with all updates I’m able to to share publicly. Stay tuned!

Thesis update!


Just a quick update regarding my thesis!

Project Dreadnought is going extremely well, and I’m excited to continue this work for my Master’s thesis over the next year. This quarter marks the official beginning of production and muscle simulation tests. I’m now producing the animation collaboratively with a small team of co-op students and graduate researchers, and I’m so glad to work with such a passionate group of people!

Because I am working with so much unpublished research for this research block, I will not update my thesis page until further notice. Stay tuned!



Thesis: Annotations of Jurassic Park and The Land Before Time


ANNOTATION #1: Jurassic Park

Spielberg, Steven, dir. Jurassic Park. Universal, 1993. Film.

Jurassic Park is widely acclaimed as a classic action film with revolutionary visual effects that hold their own over 20 years later. In addition to being a high-octane Hollywood adventure, Jurassic Park is the means by which millions of viewers in the 1990s were introduced to “the modern dinosaur.” Prior to the film’s release, dinosaurs were widely perceived as cold-blooded, sluggish, dull, and lizard-like. Although Jurassic Park is by no means a documentary and contains more than its share of glaring paleontological inaccuracies, Spielberg’s efforts to work with paleontologists such as Jack Horner gave the public a more realistic view of dinosaurs as warm-blooded, active and birdlike. As such, it is an important film to study for a baseline understanding of paleoart in a cinematic context.

While watching Jurassic Park this time around, I focused on the scenes featuring Brachiosaurus, a large sauropod (pictured above). Brachiosaurs are proportioned differently than titanosaurs and have much larger forelimbs and a more vertical posture, but the two species likely filled similar environmental niches and would have exhibited similar behaviors. Industrial Light and Magic (ILM)’s CG reconstructions of Brachiosaurus were among the first in the field of digital paleo art, so I learned much by analyzing the animal’s movements. It was also extremely helpful to look at how the art directors chose to frame such a massive creature, as scale is one of the defining points of this term’s Dreadnoughtus research.

The first scene of Jurassic Park to fully reveal a dinosaur featured a brachiosaurus rearing on its hind legs to reach some high-growing foliage. I watched this scene with Emma Fowler, a Paleontology / Digital Media undergraduate who is working on the Dreadnoughtus Project with me. The first thing she pointed out about Jurassic Park’s Brachiosaurus was the lack of motion in the animal’s shoulder girdle. Having to support the massive weight of this dinosaur would result in a lot of shoulder movement, especially when it crashes down to the ground after rearing up. The weight shifting of the animation was pretty inaccurate all around — not surprising, since these models were among the first to be digitally animated. The technology was incredible for the time, but studying the limitations of the past will allow us to create more realistic animation right now.

We also had several issues with the rearing action itself, which prompted a discussion about sauropod behavior. Would this animal rear up? Could it? The rearing pose in Jurassic Park was extremely off-balance, with the Brachiosaurus’ center of gravity resting behind its legs and near its tail. We determined that this dinosaur was more likely to collapse and break its spine than stick the landing, which would have made for a MUCH more entertaining opening. But even with more realistic weight shifting, would a sauropod ever really need to rear on its hind legs? It’s an impressive motion, but the main goal of this reconstruction is accuracy.

Anatomical issues aside, the cinematography of this scene is iconic and extremely effective. When revealing the dinosaur, the camera starts at waist level, then slowly pushes to the right and tilts upward to reveal the animal’s huge mass. This point of view emphasizes the size of the brachiosaurus, especially its vertical height. This shot is definitely one I’ll consider including in the reconstruction (though I’ll substitute Dr. Grant with a fleeing theropod). With any luck, including a similar shot will evoke the same feeling of spectacle that it did to viewers of Jurassic Park in 1993. We want Project Dreadnought to be a spectacle in its own right, while passively educating those who watch it with modern paleontology research.

Jurassic Park has one other sauropod scene which occurs after a high-intensity Tyrannosaurus chase. When Dr. Grant and his precocious companions are seeking shelter in a tree, they stumble across a goofy animatronic Brachiosaurus who plays tug-of-war with a tree branch and sneezes on them. I’m including this as an example of typical Hollywood sauropod behavior that dates way back to Windsor MacKay’s Gertie the Dinosaur. People tend to fixate on the fact that these animals are herbivores (or “veggiesaurs”), and depict them like big, doofy cows. Dr. Lacovara has expressed an interest in capturing aggressive and territorial sauropod behavior, so this scene is a perfect example of what I shouldn’t do. This animals weighed nearly 50 tons! Do you know what else likes eating leaves? An African bull elephant, and you wouldn’t try to poke one of those with a tree branch.



ANNOTATION #2: The Land Before Time

Bluth, Don, dir. The Land Before Time. Universal, 1988. Film.

When I asked a few faculty members for viewing lists that would fit this project, John Berton recommended I take a look at animated films. He warned that animation will have to be viewed in the abstract, but would still be of some use. I watched Windsor McKay’s Gertie the Dinosaur, Disney’s Fantasia, and a number of other animations, but settled on annotating my favorite childhood movie: The Land Before Time.

I was surprised to learn that The Land Before Time was produced by Steven Spielberg and George Lucas, but in retrospect it makes a lot of sense. Spielberg was going through his midlife dinosaur craze at the time, and would go on to make Jurassic Park immediately afterwards. Despite being an animated children’s movie, the direction was excellent —  The Land Before Time truly is a heartbreaking story with beautiful art direction, lighting and matte paintings. I hadn’t seen it in over a decade, and I was delighted by how well it held up when I watched it as an adult.

The Land Before Time stars a family of “long-necks,” or Apatasaurs. Apatasaurs were another clade of sauropod dinosaurs related to both brachiosaurs and titanosaurs. The main character Littlefoot is a hatchling and has extremely exaggerated movements and proportions, but the first act focuses a lot on his mother and grandparents. These are the characters I focused on for the purposes of this annotation, as they were SLIGHTLY more naturally proportioned, and provided some useful insights for framing, drama, and behavior.

The most dynamic scene by far is when Littlefoot’s mother fights off a “sharptooth,” a large theropod. I watched this scene with Emma Fowler as well. In a surprising twist of expectations, we determined that Littlefoot’s mother actually exhibited more realistic sauropod behavior than the dinosaurs in Jurassic Park did, albeit heavily abstracted and exaggerated. Littlefoot’s mother fought with her tail, shifting the weight through her hips whenever she moved to deliver a blow. I compared this motion to videos of extant reptiles such as crocodiles, iguanas and Komodo dragons, and the posturing is similar across the board.

Littlefoot’s mother does rear up on her hind legs at one point, but unlike in the scene from Jurassic Park, she keeps her tail raised while doing so. Her center of gravity is also positioned towards the front of her body and the motion seems much more plausible. She seems to do this out of desperation, as she is backed into a corner, instead of for feeding purposes. This coincides with the . For the rest of the fight, she raises her hind legs one at a time to build potential energy before swiping with her tail, which seems like a much more realistic way of posturing.

The cinematography of this scene was also incredible, and will be one of the main sources I draw from for the Dreadnoughtus tail-whip shots. Many of the shots are done with a telephoto lens, which lengthens the space and extends the sauropod’s already dynamic tail and neck. The camera angles are low, a traditional , and cut between close-up detail shots and dynamic, silhouetted long shots. This contrast emphasizes the scale of the dinosaurs. Even though my animation will not feature a fighting theropod, I believe that I can still pull off a dynamic narrative by using these techniques.

Thesis: Viewing List


Now that I’ve settled on a project about the cinematography of giant dinosaurs, I decided to ask a few faculty members for film recommendations. Here’s what I got In Response:

From Nick Juschyschyn: 

  • The latest Godzilla
  • Desolation of Smaug
  • Pacific Rim
  • Peter Jackson’s King Kong
  • Starship Troopers
  • Honorable Mention: Cloverfield

From John Berton:

  • Jurassic Park is certainly one of the best references in this regard. The sequels also but I know on the first one this was a really difficult problem and the onus was on the VFX team to make it work when no-one was sure that it would. In later films the pressure was off and I think they were not as precise in their solutions.
  • Animated films will also be of some use, although they need to be viewed in the abstract. Gertie the Dinosaur by Windsor McCay is The Classic. Lots of tricks of scale there.Fantasia has a great dinosaur sequence also. [Added by Val: The Land Before Time 1, directed by Steven Spielberg]
  • Other great live action or stop-mo dinosaurs are in The Lost World, King Kong (1933 and 2005), Valley of the Gwangi, I also recommend Walking with Dinosaurs  and Dinotopia.  There’s good stuff and bad stuff, but the contrast is what you are looking for. What makes the good shots work and the bad shots fail?
  • Other scale-oriented monsters that are well done are in Them!, Jason and the Argonauts, 7th Voyage of Sinbad, Clash of the Titans (1981 and 2010), Godzilla (all, even and maybe especially, the new one), Ghostbusters, Lord of the Rings/The Hobbitt (trolls and ents and stone giants), John Carter, Guardians of the Galaxy.
  • Comprehensive List: http://en.wikipedia.org/wiki/List_of_giant-monster_films
  • There’s lots of references out there. Watch as many as you can, bot the good and bad so that you know the difference. Also look at the real world of elephants, giraffes, shipping container cranes, monster trucks and construction equipment. Mobile launch pad assemblies, To that end, you might take a look at Koyanisqaatsi. 

I’ll be picking five of these and writing annotations this quarter, specifically focusing on cinematography that I want to use as reference for my animatic. Whenever possible I’d like to focus on the representation of sauropods, since their odd proportions will probably require very specific camera angles to capture the action. Because of this, I think I’ll be starting with Jurassic Park and The Land Before Time. This will also give me the opportunity to highlight the problems with their sauropod anatomy and posture (of which there are a-plenty) so I can avoid them in my own project.

Larger than Life: Communicating the Scale of Prehistoric CG Animals


This week I created a proposal for my 10-week research block next quarter! I’m excited to finally delve into the Dreadnoughtus project (pending approval). I decided to limit the scope of this research to one topic: Scale. I still have (roughly 60) tons of research to do before I feel comfortable making any qualified statements about Dread’s anatomy, but scale in cinema is a topic I have a working knowledge of. I want my Dreadnoughtus reconstruction to elicit the same cinematic appeal as a good monster movie, but while implicitly educating the audience with thoroughly-researched biomechanics and ecologically plausible sauropod behavior.

Below I’ve copied over the abstract of my paper and linked to a PDF. Please feel free to read and give feedback – this research will go a long way in determining the look and feel of my Master’s thesis, so comments and opinions are welcome!



Larger than Life – Communicating the Scale of Prehistoric CG Animals

Abstract — Since the earliest days of cinema, toying with the perception of scale has given filmmakers the ability to create spectacular creatures that could never exist in the physical world. With the flexibility of CG visual effects, this trend has persisted in the modern day, and blockbuster movies featuring enormous monsters are just as popular as ever. The trend of scaling creatures to impossible proportions for dramatic effect becomes problematic when filmmakers use this technique on non-fictional creatures. Prehistoric animals in particular have very few scientifically accurate appearances in popular culture, which means that films such as Jurassic Park play an enormous role in determining the public’s view of these animals. When filmmakers arbitrarily adjust the scale of dinosaurs to make them appear more fearsome, it can be detrimental to the widespread perception of prehistoric life on Earth.

For my Digital Media Master’s Thesis, I intend to create an accurate digital reconstruction of Dreadnoughtus schrani, a 77-million year old titanosaur with the largest calculable land mass of any terrestrial animal ever discovered. This animation will target a general audience with the intention of appearing in documentaries. I propose to spend the next quarter researching effective monster cinematography and creating an animatic that successfully conveys the massive scale of Dreadnoughtus while still maintaining scientifically accurate proportions. My goal is to prove that despite the freedom of scale in computer graphics, there are more effective ways to demonstrate massive size than simply making the subject larger than it should be.

Thesis Ideation: Reference


During my meeting with Dr. Lacovara yesterday, we traded reference materials and resources to keep in mind when going forward. Here they are!


This Puertasaur reconstruction is pretty awesome, albeit a little too jiggly for my tastes. Either way, it’s a good starting point, especially since it was all created in Maya and ZBrush! I’m willing to learn Houdini for this project, but if I’d prefer spending limited time learning new software. Dr. Lacovara specifically said that he wanted Dreadnoughtus to be a bit jiggly, since it would be a massive, fleshy mass of skin, fat, muscle, and air pockets.


As goofy as their martial art style may be, these giraffes are pretty badass. Because of the frailty of sauropod skulls and neck vertibrae, Dreadnoughtus probably wouldn’t do too much neck combat. That being said, the cinematography and behaviors in this documentary-style video are worth taking note of.


At :26, this komodo dragon tail whips a monkey. While this isn’t necessarily the best reference (since komodo dragons have splayed stances and Dreadnoughtus is upright with a wide gate), The quadrupedal reference might be useful.


This movie makes me cry every single time. Stylized as it may be, any Lucas / Spielberg cinematography depiction of a sauropod fighting with its tail should be useful!


This bull elephant is also pretty fantastic reference:

Thesis Ideation: Meeting with Dr. Lacovara, 12/7


Yesterday I had a meeting with Dr. Lacovara and Emma Fowler, during which we discussed our thoughts and hopes for the Dreadnoughtus reconstruction. We talked a bit about the audience, narrative, and purpose of the animation, and traded reference videos and shot ideas. I was excited to learn that Dr. Lacovara has been contacted by several different filmmakers looking to pitch a Dreadnoughtus documentary to The Discovery Channel, National Geographic, the BBC, and David Attenborough. Any animation that I create for my thesis will be viewed by a wide audience, thanks to the fact that Dreadnoughtus is already an internationally recognized name connected to Drexel.

I still need to formulate a specific thesis statement and research question, but Dr. Lacovara raised some very interesting points during our meeting. He expressed frustration that sauropod dinosaurs are often interpreted as sluggish, dopey and bovine — easy prey for any passing predator.  He proposed that a creature as massive and powerful as Dreadnoughtus would likely be very intimidating to theropod predators, especially considering the fossil evidence indicating that Dread’s tail was used as a weapon. Due to the unusual completeness of the fossil holotype, we can estimate with relative certainty that Dreadnoughtus was 65 tons of muscle-bound long-neck.

Lacovara suggested I instead consider the African bull elephant, a herbivore that is nonetheless deadly and incredibly territorial. Hippos also fit into this category, and are notoriously dangerous. A cow may be funny, but an angry bull is fiercome. Giraffes are objectively absurd looking, but even they exhibit impressive territorial behavior.

Continue reading

Thesis Ideation: Past Work with Dreadnoughtus

"Dreadnoughtus" by Jennifer Hall

“Dreadnoughtus” by Jennifer Hall

When I was little, I wanted to be a paleontologist more than anything in the world. The idea of discovering the fossilized remains of ancient creatures who ruled the earth millions of years before the evolution of modern mammals was just SO COOL that ten-year-old Valentina couldn’t imagine doing anything else.

Although  I ended up going into Digital Media (a field of study which I am absolutely in love with), I still made a point to keep in touch with my childhood interests. I’ve been volunteering at the Drexel Paleontology lab since the Summer after my freshman year, 3D laser scanning the fossils of Dreadnoughtus schrani, a 70-million year old sauropod dinosaur. Dreadnoughtus takes the prize for having the highest calculable mass of any land animal ever discovered, and was recently featured in just about every major news outlet ever.  It’s a pretty big (hah! Pun!) deal.

Continue reading