Tag: Interstellar

Christopher Nolan’s explosion

In May, Total Film reported that the production team of Tenet, led by director Christopher Nolan, found that using a second-hand Boeing 747 was better than recreating a scene involving an exploding plane with miniatures and CGI. I’m not clear how exactly it was better; Total Film only wrote:

“I planned to do it using miniatures and set-piece builds and a combination of visual effects and all the rest,” Nolan tells TF. However, while scouting for locations in Victorville, California, the team discovered a massive array of old planes. “We started to run the numbers… It became apparent that it would actually be more efficient to buy a real plane of the real size, and perform this sequence for real in camera, rather than build miniatures or go the CG route.”

I’m assuming that by ‘numbers’ Nolan means the finances. That is, buying and crashing a life-size airplane was more financially efficient than recreating the scene with other means. This is quite the disappointing prospect, as must be obvious, because this calculation limits itself to a narrow set of concerns, or just one as in this case – more bang for the buck – and consigns everything else to being negative externalities. Foremost on my mind is carbon emissions from transporting the vehicle, the explosion and the debris. If these costs were factored in, for example in terms of however much the carbon credits would be worth in the region where Nolan et al filmed the explosion, would the numbers have still been just as efficient? (I’m assuming, reasonably I think, that Nolan et al aren’t using carbon-capture technologies.)

However, CGI itself may not be so calorifically virtuous. I’m too lazy in this moment to cast about on the internet for estimates of how much of the American film industry’s emissions CGI accounts for. But I did find this tidbit from 2018 on Columbia University’s Earth Institute blog:

For example, movies with a budget of $50 million dollars—including such flicks as Zoolander 2, Robin Hood: Prince of Thieves, and Ted—typically produce the equivalent of around 4,000 metric tons of CO2. That’s roughly the weight of a giant sequoia tree.

A ‘green production guide’ linked there leads to a page offering an emissions calculator that doesn’t seem to account for CGI specifically; only broadly “electricity, natural gas & fuel oil, vehicle & equipment fuel use, commercial flights, charter flights, hotels & housing”. In any case, I had a close call with bitcoin-mining many years ago that alerted me to how energy-intensive seemingly straightforward computational processes could get, followed by a reminder when I worked at The Hindu – where the two computers used to render videos were located in a small room fit with its own AC, fixed at 18º C, and when they were rendering videos without any special effects, the CPUs’ fans would scream.

Today, digital artists create most CGI and special effects using graphics processing units (GPUs) – a notable exception was the black hole in Nolan’s 2014 film Interstellar, created using CPUs – and Nvidia and AMD are two of the more ‘leading’ brands from what I know (I don’t know much). One set of tests whose results a site called ‘Tom’s Hardware’ reported in May this year found an Nvidia GeForce RTX 2080 Ti FE GPU is among the bottom 10% of performers in terms of wattage for a given task – in this case 268.7 W to render fur – among the 42 options the author tested. An AMD Radeon RX 5700 XT GPU consumed nearly 80% as much for the same task, falling in the seventh decile. A bunch of users on this forum say a film like Transformers will need Nvidia Quadro and AMD Firepro GPUs; the former consumed 143 W in one fur-rendering test. (Comparability may be affected by differences in the hardware setup.) Then there’s the cooling cost.

Again, I don’t know if Nolan considered any of these issues – but I doubt that he did – when he ‘ran the numbers’ to determine what would be better: blowing up a real plane or a make-believe one. Intuition does suggest the former would be a lot more exergonic (although here, again, we’re forced to reckon with the environmental and social cost of obtaining specific metals, typically from middle-income nations, required to manufacture advanced electronics).

Cinema is a very important part of 21st century popular culture and popular culture is a very important part of how we as social, political people (as opposed to biological humans) locate ourselves in the world we’ve constructed – including being good citizens, conscientious protestors, sensitive neighbours. So constraining cinema’s remit or even imposing limits on filmmakers for the climate’s sake are ridiculous courses of action. This said, when there are options (and so many films have taught us there are always options), we have a responsibility to pick the more beneficial one while assuming the fewest externalities.

The last bit is important: the planet is a single unit and all of its objects occupants are wildly interconnected. So ‘negative externalities’ as such are more often than not trade practices crafted to simplify administrative and/or bureaucratic demands. In the broader ‘One Health’ sense, they vanish.

For space, frugality is a harmful aspiration

Ref:

‘ISRO’s Chandrayaan-2 mission to cost lesser than Hollywood movie Interstellar – here’s how they make it cost-effective’, staff, Moneycontrol, February 20, 2018. ∞

‘Chandrayaan-2 mission cheaper than Hollywood film Interstellar’, Surendra Singh, Times of India, February 20, 2018. ∞

The following statements from the Moneycontrol and Times of India articles have no meaning:

  1. The cost of ISRO’s Mars Orbiter Mission was less than the production cost of the film Gravity.
  2. The cost of ISRO’s Chandrayaan 2 mission is expected to be less than the production cost of the film Interstellar.

It’s like saying the angular momentum of a frog is lower than the speed of light. “But of course,” you’re going to say, “we’re comparing angular momentum to speed – they have different dimensions”. Well, the production cost of a film and mission costs also have different dimensions if you cared to look beyond the ‘$’ prefix. That’s because you can’t just pick up two dollar figures, decide which one’s lower and feel good about that without any social and economic context.

For example, what explains the choice of films to compare mission costs to? Is it because Gravity and Interstellar were both set in space? Is it because both films are fairly famous? Is it also because both films were released recently? Or is it because they offered convenient numbers? It’s probably the last one because there’s no reason otherwise to have picked these two films over, say, After Earth, Elysium, The Martian, Independence Day: Resurgence or Alien: Covenant – all of which were set in space AND cost less to make than Interstellar.

So I suspect it would be equally fair to say that the cost of C’yaan 2 is more than the budget of After Earth, Elysium, The Martian, Independence Day: Resurgence or Alien: Covenant. But few are going to spin it like this because of two reasons:

  1. The cost of anything has to be a rational, positive number, so saying cost(Y) is less than cost(X) would imply that cost(X) > cost(Y) ≥ 0; however, saying cost(Y) is greater than cost(X) doesn’t give us any real sense of what cost(Y) could be because it could approach ∞ or…
  2. Make cost (Y) feel like it’s gigantic, often because your reader assumes cost(Y) should be compared to cost(X) simply because you’ve done so

Now, what comparing C’yaan 2’s cost to that of making Interstellar achieves very well is a sense of the magnitude of the number involved. It’s an excellent associative mnemonic that will likely ensure you don’t forget how much C’yaan 2 cost – except you’d also have to know how much Interstellar cost. Without this bit of the statement, you have one equation and two variables, a.k.a. an unsolvable problem.

Additionally, journalists don’t use such comparisons in other beats. For example, when the Union budget was announced on February 1 this year, nobody was comparing anything to the production costs of assets that had a high cultural cachet. Rs 12.5 crore was Rs 12.5 crore; it was not framed as “India spends less on annual scholarships for students with disabilities than it cost to make Kabali“.

This suggests that such comparisons are reserved by some journalists for matters of space, which in turn raises the possibility that those journalists, and their bosses, organisations and readers, are prompted to think of costs in the space sector as something that must always be brought down. This is where this belief becomes pernicious: it assumes a life of its own. It shouldn’t. Lowering costs becomes a priority only after scientists and engineers have checked tens, possibly hundreds, of other boxes. Using only dollar figures to represent this effort mischaracterises it as simply being an exercise in cost reduction.

So, (risking repetition:) comparing a mission cost to a movie budget tells us absolutely nothing of meaning or value. Thanks to how Moneycontrol’s phrased it, all I know now is that C’yaan 2 is going to cost less than $165 million to make. Why not just say that and walk away? (While one could compare $165 million to mission costs at other space agencies, ISRO chief K. Sivan has advised against it; if one wants to compare it to other PSUs in India, I would advise against it.) The need to bring Interstellar into this, of course, is because we’ve got to show up the West.

And once we’re done showing up the West, we still have to keep. Showing up. The West. Because we’re obsessed with what white people do in first-world countries. If we didn’t have them to show up, who knows, we’d have framed ISRO news differently already because we’d have been able to see $165 million for what it is: a dimensionless number beyond the ‘$’ prefix. Without any other details about C’yaan 2 itself, it’s pretty fucking meaningless.

Please don’t celebrate frugality. It’s an unbecoming tag for any space programme. ISRO may have been successful in keeping costs down but, in the long run, the numbers will definitely go up. Frugality is a harmful aspiration vis-à-vis a sector banking on reliability and redundancy. And for fuck’s sake, never compare: the act of it creates just the wrong ideas about what space agencies are doing, what they’re supposed to be doing and how they’re doing it. For example, consider Sivan’s answer when asked by a Times of India reporter as to how ISRO kept its costs down:

Simplifying the system, miniaturising the complex big system, strict quality control and maximising output from a product, make the missions of Indian space agency cost-effective. We keep strict vigil on each and every stage of development of a spacecraft or a rocket and, therefore, we are able to avoid wastage of products, which helps us minimise the mission cost.

If I didn’t know Sivan was saying this, I’d have thought it was techno-managerial babble from Dilbert (maybe with the exception of QC). More importantly, Sivan doesn’t say here what ISRO is doing differently from other space agencies (such as, say, accessing cheaper labour), which is what would matter when you’re rearing to go “neener neener” at NASA/ESA, but sticks to talking about what everyone already does. Do you think NASA and ESA waste products? Do they not remain vigilant during each and every stage of development? Do they not have robust QC standards and enforcement regimes?

Notice here that Sivan isn’t saying “we’re doing it cheaper than others”, only that doing these things keeps the space agency “cost-effective”. Cost-effective is not the same as frugal.

Featured image: The Moon impact probe that went up on the PSLV C11 mission along with Chandrayaan 1. Credit: ISRO.

All the science in ‘The Cloverfield Paradox’

I watched The Cloverfield Paradox last night, the horror film that Paramount pictures had dumped with Netflix and which was then released by Netflix on February 4. It’s a dumb production: unlike H.R. Giger’s existential, visceral horrors that I so admire, The Cloverfield Paradox is all about things going bump in the dark. But what sets these things off in the film is quite interesting: a particle accelerator. However, given how bad the film was, the screenwriter seems to have used this device simply as a plot device, nothing else.

The particle accelerator is called Shepard. We don’t know what particles it’s accelerating or up to what centre-of-mass collision energy. However, the film’s premise rests on the possibility that a particle accelerator can open up windows into other dimensions. The Cloverfield Paradox needs this because, according to its story, Earth has run out of energy sources in 2028 and countries are threatening ground invasions for the last of the oil, so scientists assemble a giant particle accelerator in space to tap into energy sources in other dimensions.

Considering 2028 is only a decade from now – when the Sun will still be shining bright as ever in the sky – and renewable sources of energy aren’t even being discussed, the movie segues from sci-fi into fantasy right there.

Anyway, the idea that a particle accelerator can open up ‘portals’ into other dimensions isn’t new nor entirely silly. Broadly, an accelerator’s purpose is founded on three concepts: the special theory of relativity (SR), particle decay and the wavefunction of quantum mechanics.

According to SR, mass and energy can transform into each other as well as that objects moving closer to the speed of light will become more massive, thus more energetic. Particle decay is what happens when a heavier subatomic particle decomposes into groups of lighter particles because it’s unstable. Put these two ideas together and you have a part of the answer: accelerators accelerate particles to extremely high velocities, the particles become more massive, ergo more energetic, and the excess energy condenses out at some point as other particles.

Next, in quantum mechanics, the wavefunction is a mathematical function: when you solve it based on what information you have available, the answer spit out by one kind of the function gives the probability that a particular particle exists at some point in the spacetime continuum. It’s called a wavefunction because the function describes a wave, and like all waves, this one also has a wavelength and an amplitude. However, the wavelength here describes the distance across which the particle will manifest. Because energy is directly proportional to frequency (E = h × ν; h is Planck’s constant) and frequency is inversely proportional to the wavelength, energy is inversely proportional to wavelength. So the more the energy a particle accelerator achieves, the smaller the part of spacetime the particles will have a chance of probing.

Spoilers ahead

SR, particle decay and the properties of the wavefunction together imply that if the Shepard is able to achieve a suitably high energy of acceleration, it will be able to touch upon an exceedingly small part of spacetime. But why, as it happens in The Cloverfield Paradox, would this open a window into another universe?

Spoilers end

Instead of directly offering a peek into alternate universes, a very-high-energy particle accelerator could offer a peek into higher dimensions. According to some theories of physics, there are many higher dimensions even though humankind may have access only to four (three of space and one of time). The reason they should even exist is to be able to solve some conundrums that have evaded explanation. For example, according to Kaluza-Klein theory (one of the precursors of string theory), the force of gravity is so much weaker than the other three fundamental forces (strong nuclear, weak nuclear and electromagnetic) because it exists in five dimensions. So when you experience it in just four dimensions, its effects are subdued.

Where are these dimensions? Per string theory, for example, they are extremely compactified, i.e. accessible only over incredibly short distances, because they are thought to be curled up on themselves. According to Oskar Klein (one half of ‘Kaluza-Klein’, the other half being Theodore Kaluza), this region of space could be a circle of radius 10-32 m. That’s 0.00000000000000000000000000000001 m – over five quadrillion times smaller than a proton. According to CERN, which hosts the Large Hadron Collider (LHC), a particle accelerated to 10 TeV can probe a distance of 10-19 m. That’s still one trillion times larger than where the Kaluza-Klein fifth dimension is supposed to be curled up. The LHC has been able to accelerate particles to 8 TeV.

The likelihood of a particle accelerator tossing us into an alternate universe entirely is a different kind of problem. For one, we have no clue where the connections between alternate universes are nor how they can be accessed. In Nolan’s Interstellar (2014), a wormhole is discovered by the protagonist to exist inside a blackhole – a hypothesis we currently don’t have any way of verifying. Moreover, though the LHC is supposed to be able to create microscopic blackholes, they have a 0% chance of growing to possess the size or potential of Interstellar‘s Gargantua.

In all, The Cloverfield Paradox is a waste of time. In the 2016 film Spectral – also released by Netflix – the science is overwrought, stretched beyond its possibilities, but still stays close to the basic principles. For example, the antagonists in Spectral are creatures made entirely as Bose-Einstein condensates. How this was even achieved boggles the mind, but the creatures have the same physical properties that the condensates do. In The Cloverfield Paradox, however, the accelerator is a convenient insertion into a bland story, an abuse of the opportunities that physics of this complexity offers. The writers might as well have said all the characters blinked and found themselves in a different universe.