Do the ballooning lifeforms ('ballonts') of Avatar 3 make sense?

By Gert van Dijk and Abbydon 

If you like speculative biology, you cannot escape the Avatar films: they are spectacular. Regular readers will know that this blog likes its science 'well done' rather than 'medium' or 'rare'. But with television and film 'medium' is usually the best you can hope for. If the story is good enough, we are willing to suspend disbelief. The Avatar films are spectacular but have their share of biomechanical problems: the illogical anatomy and gaits of Avatar's six-legged animals were something best ignored in the first film, and the skimwing's size and mode of swimming in the second film did not withstand close inspection either. This post is aimed at the third film ('Fire and Ash'); as that is not even out yet, isn't it too early to start dissecting its biology? 

Based on the trailer, we thought we could have a first close look. 'We' here means Abbydon, who is a physicist, and me (Gert van Dijk/ Sigmund Nastrazzurro). Abbydon has his own blog and has written guest posts here before, on the subject of aerographene and foam as a way to make viable 'ballonts'. 'Ballonts', by the way, is a term one of us (Gert) came up with to describe life forms that move through the air using a lighter-than-air principle. At one point, I imagined a large array of floating lifeforms on Furaha, ranging from tiny aeroplankton to immense 'zeppeloons'. That bubble burst when I did the mathematics that proved that small ballonts simply could not work on an Earth-like planet, so all those lifeforms underwent a sad but sudden mass extinction. If you wish to follow the mathematics (just Archimedes' Principle, really), there is a list of posts at the end of this post. 

Click to enlarge; source: Avatar 3 trailer

The trailer for Avatar 3 is out, and it's got ballonts in it. Seeing that nature seems to conspire against ballonts, we looked at it critically. Let's start with a description. 

Click to enlarge; from Avatar trailer

There seem to be two ballont species: a large one, a 'barge', towed by a smaller one, the 'tug'. Apparently, these are known as 'medusa' and 'manta kite', respectively. A Na'avi-made ship is suspended from the barge animal so the Na'avi can use it for aerial transport. The medusa/barge animal largely consists of a large sac, elongated from front to back. It has two lateral vertical surfaces that we will call sails. Tendrils hang down and move about a bit; these are probably there to feed with and to anchor the animal. The tug is much smaller and has undulating fins, rather like Earth's rays, cuttlefish and Furahan cloakfish. Those fins propel it. 

What does this tell us? 

Ballonts need to be very large on an Earth-like planet to work (read the posts on ballonts to understand why). Gravity on the moon Pandora, where all of this takes place, is said to be low, which sounds good for balloons. But, and this may surprise you, low gravity doesn't make a balloon more practical! 'Practical', as far as a balloon goes, means a small bladder and a large liftable mass. On Earth, physical circumstances makes balloons impractical by dictating that they must have a very, very large bladder to lift even a small mass. Gravity does NOT influence the balance between the size of the bladder and the mass to be lifted, and so does not help to make a balloon more practical. Two things that do help are a high density of the atmosphere, which can be achieved by adding heavy gases to it, and a high pressure. Pandora's atmosphere is said to have a density that is 20% more than that of Earth, while the surface pressure is a bit lower at 0.9 atmosphere. Those changes are not impressive from a ballooning point of view. 

The Pandoran barge looks very large, which it has to be; so far so good. But why does it have those two large sails at its sides? To catch the wind for propulsion? We hope not, as that cannot work! Balloons are, by their nature, as light as the air around them, so they will, after a short while, move at exactly the same speed as the air around them. That leaves no wind to power anything! You can only harness the power of the wind if the air moves relative to you, for instance because you are held back by the ground or by water. 

Or do the 'sails' serve some other purpose? Are they themselves a source of propulsion? They could perhaps function like oars, folded up when moving forwards and spread out when going backwards. Or do they undulate? As they are vertical, undulation would allow vertical but not horizontal mobility. But the sails look completely immobile in the trailer. The barges do not seem to have any kind of propulsion mechanism, and if they did, they probably wouldn't have to be towed. Do the sails serve another purpose, such as heating? This is unlikely, as they are transparent; the sac should offer enough surface area anyway. Do they then help to orientate the animal with help of the wind, for instance when the animal is tethered (if it can do that)? For orientation you would want them at one end of the animal, not the middle. In short, we cannot make any sense of the barge's sails. 

Is the tug, the manta kite, large enough to float? Without a better estimate of its size, there is no way to check. The undulating fins can provide some propulsion force in air, but probably not much. If this were an animal swimming in water, fins of this relative size would work because they would displace a substantial volume of water, which is heavy. But swimming through air differs from swimming through water in various ways: there is about a thousandfold difference in density that affects thrust and drag, as well as a fiftyfold difference in viscosity. No air animal use undulation to achieve true flight on Earth, making it difficult to predict how well undulating flight would work out. Based on the low density of air, we suspect that you would need either very large or very fast-moving fins to effectively swim through air. So, whether an animal like the manta kite would swim well in air is as yet uncertain, but its proportions suggest that the animal might feel more at home under water than in the air. 

The tug does not only have to move itself but also has to drag the barge along. And 'drag' is a key word here, as in movement studies 'drag' also indicates resistance to movement. We can be certain about one thing: those immense barge sails will function as pretty efficient air brakes, making the tug's job that much harder... 

Mind you, there are some interesting loose ends about balloons and their steering that may deserve another post. Meanwhile, we hope that the film will solve the riddles. Our biggest surprise was that the trailer seems to show sails to catch the wind on a free-floating balloon; but surely the designers wouldn't have done that? 

 ---------------------------------------------------------------------- 

Posts on ballonts that help understand the physics

Ballonts III: basic physics
Ballonts IV: effects of density and pressure
Ballonts V: ballonts in gas giants
Ballonts VI: effects of the envelope 
Ballonts VIII: foam 
Ballonts IX: aerographene  

 

 

Photorealistic spidrid animation (maybe...)

 Now that The Book is out of my hands, I can pick up some old Furaha projects that I had to pause earlier on. I used Matlab, a programming and analysis tool, to produce animations to see how various animals moved. The resulting 3D representations helped a great deal to get a better feel for the gist of the movement and were very useful to get the paintings right. However, I also wanted to produce more lifelike, even photorealistic, animations. Those with good memory may remember that I used to do entire Furahan scenes using a 3D rendering programme called Vue (or Vue Infinite) from the firm E-on. There are some on my rather inactive YouTube channel.  

Producing even short animations proved to be extremely time-consuming. Vue is a raytracing programme that produces very good atmospheric results, but each frame took ages. I used to leave my PC running for one or more nights. The problem with animations is then that you only get to judge the the quality afterwards, and you may then have to do it all over again. 

The main reason I stopped making visually rich animations was that they did not contribute to The Book; by consuming time, they in fact postponed its completion. Another important reason was that programming animations was often frustrating. I could achieve my goals in Matlab, where the problems boiled down to defining rotations and translation of every body part of a radially symmetric eight-legged spidrid walking on uneven ground, with the body attitude compensating for the slope. That was complex in that it was difficult to keep track of everything, but the key parts required fairly basic trigonometry. The seriously frustrating part was to reinterpret the resulting data with Python code to direct Vue. The coordinate systems never really matched up, even if I started with both the y-axes up in both systems and similar handedness orientations of axes (that means whether the positive x,- y- and z-axes pointed in the same direction in both programmes. Even though I made sure of that, I still had to swap y- and z-axes, which in turns messed up all rotations.  

Here is an old example I posted back in 2013. There was a problem with the legs: sometimes the leg segment closest to the body (that wasn't shown) flipped around, so the pale underside is suddenly placed the wrong way. This happened when that segment rotated beyond the vertical position; let's say the rotation angles moved from 85 to 95 degrees. This was almost certainly because a function such as arc tangent interpreted 95 degrees as 5 degrees. At the time, I gave up and shelved the programme. 

I have now tried again, helped by the fact that I now have the latest and much more stable version of Vue. Sadly, Vue's parent firm (Bentley) decided that it would no longer develop Vue at all, so the 2023 version is the last one ever. The good news is that Vue is now available completely free. This version is a pleasure to work with. You do not need to be able to code at all to use it.

Here is a Matlab example of expanded spidrid animation functionality. You can see that the spidrid follows the terrain perfectly, meaning that the body posture echoes the terrain under the body precisely. You can also see something new, present in my spidrid paintings, but not yet in earlier spidrid animations. In earlier animations, the body consisted of one part whereas theyere should be two: a bottom part (the abdomen) that bears the spidrid's eight legs and a top part (the cupola or cephalothorax). The cupola can move on the abdomen. In this animation the copula is stuck firmly to the abdomen. 

In this second example, the abdomen is partly stabilised and no longer follows the terrain completely, so it stays more horizontal than in the previous animation. The cupola now moves independently and has its own tilt-dampening stabilising reflexes, so it stays even more horizontal than the abdomen. That should help the animal get a more stable platform for its senses.

After rewriting the python code the next step was raytracing with Vue. After suffering much misery and new dents in my wall where I banged my head, I got it working. The next phases include adding an abdomen and a cephalothorax, perhaps even feelers, and then it is time to add plants swaying gently in a breeze. 

Wish me patience...

---------------------------------------------------------

For those who wish to use Vue, or its associated plant editing software 'PlantFactory', you can download them here. 

As I said, this last version is quite stable. With Vue it is not that difficult to produce an image of a forest with prehistoric trees or something like that. If you wish to have more control over Vue's myriad options, be advised that this is a complex piece of kit that needs attention and time. 

For tutorials, there are several; GeekAtPlay provides good ones. 

Back from Dinocon 2025 (Exeter, UK)

Dinocon is a dinosaur-themed convention and a direct successor to the TetZooCon series of conventions that I wrote about on this blog before, in 2022 and 2018. Dinocon was a two-day conference in Exeter. It was excellent, but I will not discuss all events and talks at length; this is a blog about speculative biology, after all, not aimed directly at dinosaurs. 

Even so, I guess that most people interested in speculative biology will also be interested in dinosaurs because of the large overlap: first, dinosaurs are strange and intriguing animals, and second, much about them is speculation. Of course much can be told from their remains, which thse days sometimes even includes colours. But the gap between facts and what dinosaurs really looked like and how they behaved is huge. Every dinosaur restoration is filled with educated speculation, which is still speculation. I find palaeoart and the history of palaeoart very interesting, not only because of the artistic component, but also because the choices made in depicting an animal tell a great deal about underlying concepts. 

Dinocon, as the TetZooCon successor, fortunately still cares very much about palaeoart. Speculative biology was not prominent in the official programme, which is obviously of some concern to me. Of course, focusing squarely on dinosaurs may be the perfect choice for a convention that is just starting. I do hope that speculative biology will get more attention in future Dinocons; we'll see. Next year's version, by the way, will probably take place in early August 2026 at an as yet unknown place in the English Midlands; Birmingham was mentioned. 

Speculative biology was certainly present unofficially at Dinocon 2025. Matt Wedel, known among other things from the 'Sauropod vertebra picture of the Week' (SVPOW) blog, gave a truly excellent lecture about why the general sauropod body scheme seemed to represent a package of features that, once arrived at, seems destined to remain unaltered. He used two examples from speculative biology to illustrate his reasoning, with Dougal Dixon's 'turtosaur' as a acceptable sauropod development. 

Here's a video showing Wedel drawing attention to Dougal's 'New Dinosaurs'. 

Moreover, both C.M. Kösemen (Snaiad) and Dougal had stands at which they sold artworks and books respectively. Dougal's new version of 'The new Dinosaurs' sold out quickly. Dougal told me that the images are the same as before but that the text has been thoroughly revised. You may read a review on this new version on SVPOW, with an image of a turtosaur, right here. Andy Frasier, of 'Dragons of Wales' (here and here), had a stand at which he sold his palaeoart. Another stand sold copies of a glossy magazine, called 'Almost Real', about speculative biology that was unknown to me. I intend to write about it at a later date. 

Click to enlarge

 Here is a scene at the art show with, from left to right, Mehmet Kösemen, me, and Biblaridion.   

Something of definite interest for lovers of palaeoart and of possible interest for lovers of speculative biology is another convention about dinosaurs and palaeoart, this time in Strasbourg in France: it is called Dinoël, a combination of 'dino' and 'Noël' (Christmas). It was held for the first time in 2024, and there will be another one in 2025. Various people involved with Dinoël told me that there are no firm plans to include speculative biology yet, but they certainly did not exclude the possibility. (I did not find a website of Dinoël, but you can find it on Instagram.) 

And that brings me to the last subject of this report. There was a nice panel discussion about AI and palaeoart. As the products of AI are essentially reiterations of work produced by humans, AI 'art' is squarely based on human creativity. But if you start thinking about using art made by others, without their permission, without even referring to the source, without reimbursement, and in likely breach of copyright, the whole of AI art evokes very unpleasant associations. Darren Naish wondered out loud whether the use of AI art is in fact immoral. This is a good point, with which I agree. People who use it should think twice about the morality of using AI art.  

To end on a more upbeat note, here is a short (and jumpy; sorry) video showing a tiny bit of Exeter cathedral. I could not help noticing that some of the grotesques on the facade are very nice examples of creature design.  

And on an even more upbeat note, here is my wife, Roelien Bastiaanse, having mixed feelings about dinosaur reenactments.