Follow the reluctant adventures in the life of a Welsh astrophysicist sent around the world for some reason, wherein I photograph potatoes and destroy galaxies in the name of science. And don't forget about my website,

Saturday, 29 July 2017

Ultra Diffuse Galaxies : Revenge Of The Ghosts

Why is everyone wearing badly-ripped trousers lately ? Like, literally almost everyone, men and women alike ? If it's supposed to look sexy, it doesn't. It just looks daft. Just like when everyone in the entire world was wearing ponchos a few years ago for no good reason other than that everyone else was wearing ponchos.

Scientists aren't really know for being particularly fashion-conscious and those that do tend to go to bizarre and questionable though strangely wonderful extremes... at least when it comes to clothing. As far as research goes, however, it's a bit of another matter. There's generally almost always someone working on just about anything, but every so often something happens that transforms the dull, plodding research in some particular field into a crazy festival that's overtaken by screaming hordes of sexy young people (and often screaming hordes of drunken academics as well).

Currently the winner of the sexiest, most fashionable topic in extragalactic astronomy is undoubtedly the analysis of so-called "ultra-diffuse galaxies". You might remember I wrote about these almost two years ago when they first started hitting the headlines. For a topic this large, it's usually not worth paying too much attention to individual press releases. But two years is enough time that many, many more studies have been done, so it seems like it might be worth an update. First, some background history.

Bright Galaxies Are For Losers

For a long time it was thought that all galaxies had the same "surface brightness" - that is, they seemed to emit the same amount of light per unit area. This was a bit surprising, but then along came the discovery of Malin 1.

Malin 1 isn't a particularly faint galaxy, but it's much larger than other galaxies of the same brightness. A trickle of other giant "low surface brightness" galaxies followed over the years, but none quite so dramatic. A few people thought it might just be a selection effect - we were only seeing galaxies that were so bright that they were easy to detect, and that only much more sensitive observations would reveal greater numbers of these LSB galaxies. Other people thought that these LSB objects were just exotica - sure, Malin 1 might be super interesting, but not all that important in the grand scheme of things. Kindof like the giant squid : sure it's cool that there's an enormous version of the regular squid, but it lives in the deep ocean and literally no-one thinks about it on a daily basis except for hungry sperm whales and, presumably, other giant squid.

And daring sea captains, of course.
While the giant ghostly galaxies were sometimes viewed as nothing more than a curiosity, it was a different story for LSB dwarf galaxies. Now, to you or I dwarf might mean something that's physically small, and giant something that's physically large. Not so in astronomy, where the terms don't have strict meanings but are generally used to refer to brightness, not size. If astronomers ran hardware shops, they'd call 20W bulbs "dwarfs". That's the kind of linguistic logic we're dealing with here : no wonder we have such a stupid definition of "planet".

Anyway, like flies on a windscreen, LSB dwarf galaxies were found in large numbers relatively quickly. By and large, dwarf galaxies do tend to be smaller than giant ones, and they're also much more common. This means you don't need a large field of view to discover them. What you need are two things : 1) Long exposure times (nothing difficult about that) and 2) a method to estimate their distances. You need the distance to work out their true size, mass, and brightness, otherwise you have the classic problem of not knowing whether your cows galaxies are small or far away.

Directly measuring distances is hard, but there's a trick that makes things much easier - look for the faint galaxies in known clusters where the distance has already been measured using bright galaxies. If you see more of them in the cluster and less in a control field, you can be pretty sure that most of your discoveries will be at a similar distance. Of course, not every object will be a cluster member, and even those that are won't be at exactly the same distance - but it's a reliable enough method that your conclusions will be statistically valid.

LSB dwarves were so easy to find that they moved from the realm of "neat" to "dull" pretty quickly... at least in terms of discoveries. They became routine. They didn't stop being interesting, but... where were all the giant LSB galaxies ? Eh ?

Moaaar Galaxies ! More I Say ! Hang Those Who Talk Of Less !

That answer only came much more recently, with hundreds upon hundreds of giant LSB galaxies being discovered using more sensitive telescopes with the wide fields of view needed to detect the giant objects. Far from being exotica, they seem to be extremely common. Some clusters contain hundreds of the ghostly blighters, although in other cases their numbers are far less.

From one of the first discovery papers.
From another early paper of the heady days of 2015.

Finally some more interesting-looking objects from a more recent paper.
As you can see, most of these galaxies aren't visually very appealing (though there are some interesting exceptions). Mostly they're smooth, boring looking objects. They don't seem to have much in the way of gas or star formation - they're just red, dead pustules of stars floating through the universe like long-dead corpses, or something.

Almost all of these galaxies have been discovered using the same trick as for the LSB dwarf galaxies : looking for faint smudges that seem to be more common around clusters of known distance. We've also now got a few more accurate distance estimates, so we're pretty confident that the majority of these discoveries really are giant galaxies far away. Of course, no doubt a few of them are small objects nearby, but the majority are definitely a long way off.

Make no mistake : this is really cool. Hundreds and hundreds of giant, faint galaxies we never knew existed ! But... what is the nature of these objects ? How do they form ? Are they just like regular galaxies except that they never formed as many stars because they got bored, or did something happen to them to stunt their emotional growth ?

The Nature Of The Beast

Galaxies generally have several components : stars, gas, dust, and dark matter*. Technically all of these are optional, but usually ya gotta have dark matter. Dark matter dominates the mass of most galaxies. Gas and dust are completely optional : we know of loads of cases where galaxies just have stars and dark matter. They're really, really boring, but most of the Universe is so damn complicated we should be frickin' grateful for a little boredom once in a while.

* To keep this short I'm just gonna assume dark matter is real; for more details see these posts.

So frickin' boring.
Dust rarely (if ever) makes up much of the mass of a galaxy, but the gas can be significant - sometimes there can be far more gas than stars. This raises the question of how extreme these "dim galaxies" can get... instead of low surface brightness, could we have galaxies of no surface brightness with just dark matter and maybe some gas ?

Perhaps. That's a hugely controversial topic, which the enthusiastic or drunken reader can examine here or here. But it also makes us re-examine that definition of "dwarf". Leaving aside the gas and dust, we can measure galaxies in terms of their size, mass, and brightness. Traditionally we've been able to use brightness as a pretty decent proxy for the other two, since there seems to be a nice correlation. So calling a dim galaxy a dwarf wasn't totally mad. But these new "ultra diffuse galaxies" are very different to almost all other galaxies we've seen before : as large as the Milky Way but a hundred or a thousand times fainter.

What we'd really like to know is their total mass, i.e. how much dark matter they have. Then the question becomes : are these huge dwarves (big but not very massive) or ghostly giants (big and massive but very dim), a.k.a. "failed galaxies" ?

If you stuck a bicycle pump up Gimli's backside and applied sufficient pressure, you'd get a huge dwarf. His mass wouldn't change, but his size would.
For cosmological models, it's their total mass which is the most interesting thing. Such models traditionally only use dark matter because it makes the calculations much easier, which means they have limited predictive power when it comes to things like brightness and size. But mass they ought to be able to get right. They've been plagued with problems for the lowest-mass galaxies, predicting far more tiddly little dwarf galaxies than are actually observed, but it's looking increasingly likely that this was just because they didn't include the complicated physics of the gas and stars. So there could be plenty of small galaxies which just have too few stars to be detectable.

But these models haven't had a problem when it comes to massive galaxies : they get their numbers about right. If these new galaxies are low mass, then they could help solve the missing galaxy problem (or at least not make it any worse). If they're high mass, then...

Unlike in popular 80's movies, a few giant ghosts wouldn't spell disaster. That could just mean that we don't properly understand how gas forms stars, which we know we don't understand anyway, or how gas gets into the dark matter clumps (which we also don't know anyway). So they could even help us understand the missing midget galaxy problem. But if all these giant ghosts are massive ? Well then we've got a real problem.

Who Ya Gonna Call ? Some Dorky Astronomers, Probably, Because This Is An Astronomy Problem

Following this field over the last couple of years has been quite exciting. Sometimes the evidence swings one way, sometimes another. Currently, the balance of probabilities is shifting towards most of new giants being huge dwarves than giant scary ghosts... but the dwarves aren't having it all their own way.

Normally we can measure the total mass of a galaxy by seeing how fast it's rotating, or how fast the stars are buzzing around if they're not in rotation. We can do this by measuring the stars, but it's generally reckoned to be more accurate to use the gas. Gas is normally more extended than the stars so it lets us measure how fast the galaxies are rotating at greater distances from their centres. Unfortunately, most of these new ones don't seem to have much or any gas, so we have no choice but to resort to the stars. And that sucks, because these things are so bloody faint that measuring the speed of the stars is abominably hard. So two years later, we don't have that many mass estimates.

But we do have some. The first measurement was of a pathetic galaxy known as VCC 1287. There had been speculation that ultra-diffuse galaxies in clusters would need particularly high dark matter contents to survive being ripped apart by the other galaxies. These measurements confirmed that : this object has far more dark matter than other galaxies of a similar mass. But not so much that it could be classed as a giant, which is quite satisfying.

It's either an interesting galaxy or someone sneezed at the wrong moment...
To get around the problems of measuring this miserably faint smudge, the authors (a certain Michael Beasley et al.) used a neat trick - they measured its globular star clusters instead. Globular clusters are known to surround many galaxies in a sort of cloud. Because they contain large numbers of stars in a small area, they have higher surface brightnesses than the rest of the galaxy, making them much easier to measure. So the authors were able to use these to directly estimate how fast the galaxy is rotating*.

* Strictly speaking it's not rotating, it's swarming. The point is to know how fast the stars are moving. Whether they're rotating nicely or moving in crazy random orbits, the method is the same : speed => total mass.

But they were also able to do another clever trick. Using a known relation from other galaxies, they were able to estimate the total mass just by (in effect) counting the number of globular clusters. And they got a result that was in great agreement with their direct measurement. Hurrah !

The grey circles are VCC 1287; all the rest are a bunch of other crappy galaxies. Total mass in on the y-axis, stellar mass on the x-axis. See how VCC 1287 has way more total mass than most galaxies with that many stars ? We'll come back to that later.
Now, a bit of caution is needed because that globular cluster versus total mass relation wasn't previously calibrated for galaxies like this. But it's in such good agreement that it suggests that we can - tentatively - use it to explore the masses of other UDGs. And we don't even have to do the tricky work of measuring their speed (which is complicated), we can just count them (which is way easier). So subsequently they repeated this neat trick on another ultra-diffuse galaxy, and found that that one also is probably a huge dwarf instead of a giant ghost.

Thus far most of the UDGs we've see have been the Brienne of Tarth of the galaxy world : very interesting, but not terribly attractive. One of the rare things that astronomers agree on is that without star formation, galaxies quickly settle into the photogenically uninteresting "red and dead" phase. There are several reasons for this. First, stars are so small that they rarely collide or interact with each other except through gravity, so a collection of nothing but stars is quite well approximated by a swarm of angry bees. There's nothing to make the bees stick together, so they form something quite smooth and uninteresting to look at unless it's chasing you because you raided its hive... the analogy probably breaks down at that point.

Second, blue stars are short lived. So once star formation stops, the blue stars quickly die off and the galaxy becomes a boring homogenous red colour. And for star formation you need gas... which also happens to make things more interesting for another reason. Let's switch from bees to cats. A swarm of angry cats is pretty well randomised, but imagine a swarm of cats following catnip : much more interesting. Gas can collide with itself and stick together to form interesting structures, and within those denser structures stars can form. Gas, essentially, makes galaxies beautiful. Maybe someone should tell the make-up industry.

The first hints of UDgs with gas came indirectly, with the discovery that some of them were blue and had complex structure :

The authors of that study found that the redder ones had less stars and were overall rounder. And the bluer ones seemed to be more common in the outskirts of galaxy groups. So perhaps as they fall into the group and get harassed by the other members, they can lose gas (and some stars) and transform into the boring red objects that dominate in most cases. They also found that the UDGs were found in similar locations to the normal dwarves, suggesting that maybe these UDGs were over-inflated dwarf galaxies rather than giant ghosts.

What we'd actually like to test this is a direct measurement of the gas, not just inferring its presence by their colours. That would also let us directly measure their total mass.

Gas for some... miniature American flags for others ?

Cometh the hour, cometh the archival data. The huge ALFALFA survey is covering the whole sky visible to Arecibo, looking for hydrogen gas rather than optical galaxies. With over 24,000 galaxies catalogued to date, it's an excellent resource to mine for UDGs that may have been observed but not noticed. Which is exactly what the authors (led by PhD student Luke Leisman) of this study did, combining their results with the even larger optical Sloan Digital Sky Survey.  They found about a hundred UDGs hiding in the data, all with gas.

Indeed, "woohoo". Mostly. But there are some important limits. Except for three cases, the resolution of the gas observations isn't very good. This means they can't estimate the total mass as well as they would like. Normally with observations like this you can make pretty reliable corrections, but these galaxies are so faint that's not really possible.

Sometimes limits aren't necessarily bad. A more interesting restriction is that all of these sources are isolated, much more so than the previous results which were all in groups or clusters. That's a limit the authors imposed to avoid confusing their discoveries with other sources. They were also able to completely avoid the need to look for galaxies in clusters, because with hydrogen surveys one gets the distance measurement directly from the data.

What they found was that these galaxies are blue and irregular... and low mass. Although the mass is very uncertain, it seems unlikely that most of them are giant "failed galaxies" - they are indeed just over-inflated dwarves. So, galaxy formation theories saved then ?

Things are indeed getting weird

Errr, well... ummm... dunno ! We need to make a brief digression here, because these galaxies also have another especially weird property : they seem to be rotating much, much more slowly than expected. This plot wasn't in the paper, and I'm hesitant to do science by blog post, but it's the best way of illustrating it :

You can see that normal galaxies follow a nice clear trend : the faster they rotate, the higher their total mass. This is not so for totally optically dark hydrogen clouds, which rotate much more quickly than normal galaxies of the same mass. But the ALFALFA UDGs do the exact opposite, apparently rotating much more slowly than regular galaxies.

This is something I've covered in more detail here, but the main point is that measuring rotation for the hydrogen clouds is difficult - we don't really know if they're rotating or not, so we may not be making a fair comparison by plotting them together with normal galaxies. But for at least some of the ALFALFA UDGs, that's not the case. Three of their objects (the ones with the high resolution observations) really do look like they're rotating. And while we know that it's possible to create the appearance of slow rotation in galaxy clusters, these galaxies are bloody miles away from any others, so that can't be the explanation in this case.

So this result is potentially super-interesting. In decades of observations, no-one's ever seen a deviation like this before... and that's been a surprise, because it's not obvious from theoretical grounds why there should be such a neat relation between rotation speed and mass. We expect galaxies of low surface brightness to deviate, but so far they haven't. Could it be that there isn't really a neat relationship after all, and it was all some kind of weird selection effect ?

I asked Lesiman about this, and he gave a very cautious response - perhaps even a little too cautious. He says that there could be a biases in the selection procedure might mean that the rotation widths are underestimated. The above plot might not be comparing widths in the same way. Maybe the galaxies aren't really discs at all. Or we might be missing the fastest-rotating objects because of sensitivity effects. I agree with most of these, but I think the effect would be weak - even combining them all, I doubt it will change the end result.

To show just how strange the result is, it's worth noting that these pathetic objects...

... all lie slap-bang on the "Tully-Fisher" relation. Even the crappy ones which are certainly LSB dwarves (though they're not UDGs, which are larger). It's worth remembering, though, that we have hardly any measurements of the UDGs, so we have no idea where the majority lie on the Tully-Fisher plot.

But at least the UDGs are dwarves, right ?

Probably. The bulk of the evidence looks to be pointing that way, and there's at least one paper showing how dwarf galaxies could become so large in principle (whether or not that actually would happen according to the standard models is another matter). But there is some dissent. Enter the 44 galaxy.

Dun, dun, duuunnn !
I've already written a little bit about the galaxy Dragonfly 44, noting how although the press release claims it to be the first discovery of a "new class of galaxy", the reality is more complicated. Yet there's no doubt it's a very interesting object. Unlike the other objects it does seem to be a "failed galaxy", with as much dark matter as the Milky Way. And the authors were able to estimate the mass in two ways - from the velocity of the stars and the globular clusters, and by counting the globular clusters. The results are in good agreement.

... but, let's not be hasty. There's considerable margin of error in all of the estimates of total mass, and not just because of measurement uncertainties. The authors note :
We emphasise, however, that the halo abundance matching technique relies on total halo masses, and in our study the total halo mass is an extrapolation of the measured mass by a factor of ∼ 100. A more robust and less model-dependent conclusion is that the dark matter mass within r = 4.6 kpc is similar to the dark matter mass of the Milky Way within the same radius.
The difficulty is that you can only directly estimate the mass of dark matter from the visible matter. You don't really know if the dark matter cloud (or halo) extends much further, because if it did it wouldn't change the measured velocities.  The way the total amount of dark matter is estimated is by comparing it with numerical models... models which are known to have some pretty nasty problems.

So the direct mass measurement of this galaxy is consistent with the Milky Way as far as it can be measured. The visible light continues out to greater distances but it's too faint to use for measurements. If the usual relation holds, then Dragonfly 44 may indeed be a giant stupid galaxy that's failed at forming stars. This is not improbable, and that the globular cluster/velocity measurements have given results in good agreement with each other for VCC 1287 bodes well. And the authors also note :
Better constraints on the halo masses of UDGs may come from lensing studies of large samples. Intriguingly, a weak lensing map of the Coma cluster by Okabe et al. (2014) shows a 2σ peak at the location of Dragonfly 44. Peaks of similar significance have inferred masses of a few trillion solar masses, and unlike most other features in the map it is not associated with known bright galaxies or background structures.
Gravitational lensing can provide an additional, independent mass estimate of the entire halo. The evidence for this effect for Dragonfly 44 is weak, but intriguing.

Still, we should be cautious about that large extrapolation. And a recent paper has also found that galaxies experiencing gravitational forces from their neighbours can have large amounts of their mass removed without changing their size. How much of an effect this could have in a galaxy cluster like the one in which Dragonfly 44 resides isn't yet known.

Oh yay, more boring galaxies. I shall try to contain my excitement.

Not content to rest on the exciting laurels of Dragonfly 44, the van Dokkum-led team have analysed a bunch more galaxies in the same cluster. They found that although Dragonfly 44 is still the star of the show, 15 other galaxies could also be these ghostly giants, based on the number of globular clusters. And they also show that for those three cases where we can estimate the dark matter from velocity measurements as well as cluster counts, the galaxies lie on the usual relation. That suggests the cluster count method is reliable, even though these galaxies are so different to the usual ones.

But they also comment on the previous suggestions that most UDGs are actually huge dwarves, and here we come a bit unstuck. They note good reasons why there are differences between different studies, especially why they're finding more globular clusters than everyone else - but they don't say which method is better. One important detail is that the number of globular clusters has to be corrected for background sources that aren't associated with the galaxies, but it's not at all clear who has the better method. So I think they're jumping the gun when they claim, "The results presented here... put to rest the suggestion that most cluster UDGs are directly related to smaller galaxies of the same total luminosity." Pretty strong words*, and not really justified as yet given all the other evidence, in my opinion.

* A direct translation from academic to normal speak is difficult but it implies something like, "everyone else is wrong, shut up already."

The other worry is that it seems to be van Dokkum Against The World. It seems to be only this group that's claiming these are failed galaxies : everyone else seems quite convinced that they're inflated dwarves. It's far too early to use this as evidence against them (and I should mention that if van Dokkum is right then it supports my own research, so let's no-one cry out "bias" please because it wouldn't make any sense), but it would be nice to have an independent claim for the giant ghosts.

Late one night, two teams of astronomers are locked in a house haunted by a particularly angry dwarf galaxy.


Two years on, and it might feel as though we don't know that much more than when we started. Certainly the main issue - how massive all these new galaxies are - remains controversial, and could easily take another two years or more to settle. Considering how incredibly faint these galaxies are and the pressures of getting telescope time, it's really not so surprising that this is proving difficult.

But even if the full implications of these new discoveries have yet to be worked out, we are learning things :

  • We know that giant ghostly galaxies are not the rarities we once thought they were, they're very common.
  • We know they're found in all galaxy environments, from isolated objects up to rich clusters. We also think that some clusters are richer in UDGs than others.
  • Although UDGs have far less stars than normal galaxies of the same size, they apparently have even more globular clusters. Whatever stops them forming stars in their main body doesn't stop them forming clusters !
  • The gas content of the galaxies is highly variable. Malin 1 has the highest known gas content of any galaxy ever (it's got like, literally tonnes of the stuff), whereas most of the others seem to have no gas at all. It doesn't seem to matter how bright they are : some faint galaxies have no gas whereas others have more gas than stars.
  • UDGs come in many different shapes and sizes. Generally they have about as many stars as a large dwarf galaxy, but their size can vary considerably. Most seem to be smooth spheroids. Some are irregular. A very few, like Malin 1, are giant spirals.
So we've learned a lot, but we're still in the phase of learning more about our ignorance than anything else. We had no idea these things even existed until a couple of years ago, and now they seem to be an extremely important component of the Universe. How much of a challenge they pose to our theories is yet to be determined. 

We know of different possible formation mechanism for these objects and it's likely that all of them play some role. The question is, which is dominant ? If they're giant galaxies, then why did they form so few stars compared to other galaxies ? What mechanism affected them that didn't affect normal galaxies ? Why do some have huge amounts of gas but apparently little or no star formation whereas others are completely gas-free ? Why do some clusters have more UDGs than others which contains lots of very faint dwarf galaxies ? How much fainter can galaxies get - could some have no stars at all ?

We're only just beginning to tackle these questions, and more besides. As usual, observation is running ahead of theory. Not only does the question of whether they're huge dwarves or ghostly giants continue to cause controversy, but we also don't yet know exactly how many there are. Finding these things requires no just deep observations but also robust statistical models. Most detection methods use automatic techniques to find faint smudges in clusters that aren't found in control fields... but these automatic catalogues typically start with numbers in the millions that are selectively cut down to a few thousand candidates. This doesn't mean there are millions of undetected galaxies, but it does mean there are almost certainly a lot more undiscovered objects out there. So stay tuned, watch this space - the best is surely yet to come.

So spooookyyyy....

Monday, 24 July 2017

Arecibo Reloaded

Several years ago, as long-term readers will recall, I made a CGI model of Arecibo Observatory because my then-boss told me to. This was then turned into a laser-etched glass cube, originally just as a nice present for the Observatory employees but subsequently sold at the visitor centre. The original model took about a week to make. It was based on the original telescope schematics plus careful site-walking. It had enough details for the required glass cube, but not too much more than that.

Late last year I bought myself a VR headset. I have a review post of that in draft, but suffice to say it's rather fun - though the technology is still immature. Naturally I wanted very much to convert my own content into VR format, because when it works, it really works. The sense of immersion is much, much greater than any other format.

There are different ways to produce VR content. One is to make an interactive, game-like format, where the user can walk around a virtual environment however they like. Of course this would be the most fun sort of VR, but you lose realism and detail. Well, more accurately, if I were to try this, my result would lack realism and detail because I don't have much experience creating interactive content. I'm much more familiar with another method that can be used for VR : pre-rendered video.

Pre-rendered content means you can apply all sorts of fancy lighting and material effects that give realistic results much more easily than with the interactive approach, and you don't have to worry too much about the vertex count. I was urged to try the Unity game engine, but in the end the burden of learning an entirely new interface was just too much. Upgrading the Arecibo model to VR standard was no small task in itself; the prospect of also learning new software with a radically different approach to Blender turned, "relaxing evening" into, "yeah, after you finish work for the day, keep doing more work in the evening." So I went with what I knew : pre-rendered content in Blender.

This seems like a good opportunity to list some of the mistakes I made when learning how to get VR content working, but if you're really only here for the Arecibo stuff, feel free to skip the whole next section. What follows is a pseudo-tutorial to creating VR content; if anyone wants me to develop this into a full tutorial, then I will.

Creating VR : What Not To Do

Before I started playing with the Arecibo model I wanted to get VR content working with something simpler. The main guide I relied on was this excellent video tutorial :

I hate video tutorials and I cannot for the life of me understand why they're so popular, but this one is very good. A couple of things are worth adding/emphasising :
  • Blender's native 360 3D (spherical stereo) camera's are only possible using the Cycles rendering engine. The reason for this I'll explain in detail below.
  • Although it comes with various options for the type of VR content, e.g. top-bottom, side-by-side, these don't actually seem to do anything. You have to render the two images separately and join them together yourself (you can use Blender's sequencer for this, but you have to set things up yourself - there's no automatic, "side by side" button).
All of my content thus far has been with Blender's traditional internal rendering engine, because it's fast and I know how to use it. Rendering speed doesn't seem to be getting much emphasis these days; the focus seems to be on ever-more realistic results. Which is fine, but I don't care for waiting hours and hours for a single still image. I want speed. So Cycles, which is generally much slower than the internal engine, isn't usually for me.

Unfortunately it's not as simple as changing which rendering engine you want and re-rendering an old scene. The two engines are radically different, and this means you have to remake all the old materials in a way that Cycles can understand. Initially this seemed so daunting I decided to try and find a workaround. I'd already done some standard side-by-side 3D content, so it seemed to me that the key was to figure out how to render 360 spherical content and just do this from two different positions. What you need for the headset is an image in equirectangular format

First I tried using the "panorama" option of the camera, experimenting with the field of view setting. Although it's possible to get something sort-of reasonable with this, it's not great - the image gets very distorted at the poles. This option is best avoided.

Fortunately I came up with what I thought was a clever solution. I'd render my old scenes using the internal engine with the classic "6x 90 degree F.O.V." images. Then I'd setup a skybox as normal, but I'd use Cycles materials for each face of the box so I could then use the Cycles spherical camera. Creating a shadeless Cycles material is pretty trivial, and avoids having to learn Cycles in any real depth - and more importantly, there's no need to convert any of the old materials. Plus this would be easy to animate.

Part of a classic skybox. Each face was rendered in Blender internal using a 16 mm camera, giving a 90 degree field of view.
This actually works. When you render the above skybox (adding in the missing planes) using a Cycles panoramic, equirectangular camera, you get the following :

Seamless and perfectly distorted as an equirectangular image should be. Great ! If that's all you need - e.g. 360 degree but 2D panoramas - then there's nothing wrong with this method. You can test the images using, for example, this, or you can find whatever application you prefer for turning them into web pages. Google+ used to let you do this directly, a feature which got lost at the last update but is slowly being re-implemented.

You can also view this directly in Blender in realtime without needing a camera. What you do is to join the faces of the skybox together (make sure your textures are UV mapped), subdivide the mesh a bunch of times, then use the "to sphere" tool. That turns your skybox into a skyball, which you can view just fine in Blender's viewport.

That approach lets you skip the Cycles renderer altogether, as long as you just want a personal viewer. Another option is to use a Python script. After creating the skyball, knowing its radius and the position of each vertex, you could move each vertex to the position it should have on an equirectangular map (i.e. convert its Cartesian coordinates into spherical polar coordinates). That will get you an equirectangular map directly.

I expected that since regular 3D content just consists of two side-by-side images, I could then render two skyboxes (or skyballs) from slightly different positions, and join the two equirectangular maps together. This does not work. Don't do it !

What you get if you try this is something very strange. From one viewing angle, everything looks great on the VR headset... but as you turn your head, the sense of depth changes. Turn your head 180 degrees and you realise the sense of depth is inverted... but if you then turn your head upside-down, everything works again !

This bizarre behaviour was not at all obvious to me, and I only understood it after a lot of Google searching. It turns out that you can't ignore the fact that your eyes move in space as you turn your head. Rendering from two fixed viewpoints is not good enough - you need to account for your eyes being at slightly different physical locations depending on your viewing angle. That's why the spherical stereo mode isn't supported in Blender internal. It requires the camera to render from a different location for each horizontal pixel of the image, which the internal renderer simply doesn't support.

Technically it might be possible to write a Python script to get around this problem. But it would be ugly and incredibly slow. You'd have to render each column of pixels of the two images from different locations, accounting for the rotation of the two cameras around their common centre, and then stitch them together. Since you're going to want your images to be at least 2k on a side, that means rendering 4,000 images per frame. Don't do that. Really, your only option is to go with Cycles.

Arecibo is a complex mesh, so once I resigned myself to the need to use Cycles, initially I thought I'd start with something simpler. The ALFALFA animation seemed like a good choice : 22,000 galaxies, all with very simple, scriptable materials. That would look great in 360 3D VR, wouldn't it ? Being surrounded by a huge mass of galaxies floating past would look pretty shiny, eh ?

It would. And scripting these materials turned out to be extremely simple, which I was rather pleased with. But alas ! It didn't work. Cycles may be technically more capable than the internal render engine, but it includes an extremely irritating and hard limit of the number of image textures it supports : 1024. The only way around this is to edit the Blender source code. I'm told this would not be so difficult, but I didn't fancy trying that.

So I gave up and decided to do the thing properly. Arecibo at least didn't need a thousand different image textures.

Remodelling Arecibo Observatory

The original mesh wasn't in too bad a state. Not so long ago I'd done some tidying up to make an animation for the visitor centre, so I'd already added some details that weren't in the original model.

Of course it isn't perfect. Strange flickering besets the landscape (and to a lesser extent the trees); Blender's textures are not always as stable as they should be. The trees are rather too deciduous for the tropics, but those were the best tree sprites I could find - and overall, I rather like the forest effect.

Viewed closer, the model looks acceptable, though it lacks detail. The materials are decent, but of course they are far from perfect.

However it looks best from below. The low resolution of the landscape is a problem - this was the highest resolution available from the USGS, but it's not really enough, and manually editing it would be quite a task. And while the rocky texture looks OK from a distance, it's not so great close up. All these problems disappear from a different viewing angle.

That starts to look halfway respectable, in that the top half of the image looks respectable though the lower half not so much. Here's a reference photo for comparison.

I began the VR conversion with the existing materials. Unfortunately, I couldn't find an acceptably fast render solution with the trees, so they had to go. Much work went in to creating landscape materials that the viewer could accept as representing rocks and trees. Learning how to distribute different textures using Cycles materials was one of the hardest parts of the process, but eventually something clicked and it started to make sense.

I dallied with getting the rocky areas to have some displacement, but I couldn't get this to work well, so I stopped. Certainly there's a lot of scope for improvement, but it's fast. With the plan being to have the animation take the user on a walking-pace tour of the telescope, rendering speed was all-important.

Fortunately, the telescope itself doesn't feature too many complex materials. Converting them to Cycles format was relatively painless.

The one major exception was the main white paint material. That went through many iterations before I got the balance right. Eventually I realised two things : 1) from reference photos, the material has different levels and types of dirt depending on where it is and when the photo was taken; 2) it's actually white. Not grey - bright white. Making things a brighter shade of white in Blender is one of those things which is really very simple when you know the answer but can sometimes be unexpectedly difficult to solve : make the lights brighter ! Yes, you might then have to adjust all of your other materials, but that's what you gotta do.

To keep rendering times short, I basically disabled all of Cycles fancy lighting effects. Light bounces were reduced to their minimum values, except for transparency since I needed a few transparent materials (the fence mesh material - sometimes you can see other fences through the fence, and you need multiple "bounces" for this to render correctly). I used "branched path tracing" rather than the regular "path tracing" to make sure everything was set on minimum. That completely eliminates the grainy look that often plagues Cycles renders, reducing it back to something approaching the internal render engine in look and speed. Render times were slashed from several minutes per frame (using the default settings) down to 30 seconds.

Of course the penalty is that the render isn't as realistic as it might be. An annoyance with Cycles is that it doesn't support hemi lights, which are useful for faking diffuse background light. I had to make do with crappy sun lights instead, but beggars can't be choosers.

One important decision that had to be made was the level of detail I was prepared to add. The original was very simple - fine for distance shots, but not suitable for close-ups. Also, while the telescope schematics contain everything you could ever want about the superstructure, they contain nothing about everything else : the walkways, the waveguide, the cables - none of these are included at all. And the real telescope is, in many places, ferociously complex.

Worse, I don't have many reference pictures of some of the most complex areas - you don't tend to take photos of those places.

You can see in the above that the major girders of the telescope are all complex features made of many sub-girders and supports. I had to neglect these. It would have made the modelling process incredibly tedious and the model impossible to work with. As it was, the final vertex count was a mere 650,000 : with the girders done in full detail it would probably have been tens of millions. I suppose it might be possible to use image textures, but that can be for the next iteration. So I went with, "include all the major structures, but don't render them in full detail."

I also had to sacrifice on cables. When you're actually up there, the platform is an even messier place than it looks from the photographs. Trying to track every single cable would have been absolutely impossible.

The waveguide - which carries the signal received from the telescope back to the instruments in the control room for analysis - also had to be compromised. You can see it in the above photograph, running vertically through the image right of centre and looking a bit like an air vent. Following its precise path wasn't possible, so I simply included it where I could but it's got quite a lot of gaps in the final model. Which means my virtual model wouldn't really function. Oh well.

What I decided to try and include as much of as possible was everything else : all those other secondary details like railings, lights, boxes, signs, rivets... all those sorts of little details. Great care was taken with each reference photo to include the unique features of every part of the platform, rather than inventing random industrial details. If this is art, then it's art without much creativity in it.

The lower section rotates, of course, which is why it looks different here than in the above reference photo.

Several mistakes here. I put the stairs leading down on the inside, whereas actually they're found on the outside of the azimuth arm. The building is slightly too tall. And the two separate sections should be connected. Oh well.

Again some differences here because the arm was rotated at a different angle in the reference photos.

Of course there are a lot of difference in this dense, complex area where I had few good reference photos. Eagle-eyed viewers will notice that the dirt pattern on the diamond plate flooring changes depending on which level you're standing on.

Pretty nearly all of the details shown above - and more besides - are new for the VR video. Not everything is visible in the VR display - including, unfortunately, the signs. The signs are where I allowed myself a creative outburst.

And so, without further ado, the tour. Starting from the catwalk, it proceeds at walking pace along the top of the triangle, then descends to the central pivot section. From there, look up to see the sky as seen using Arecibo at the 21 cm wavelength. Then it jumps to the upper section of the azimuth arm and walks from one end to the other. It lasts 3 minutes. The sound is a free industrial sound I found somewhere on the web (it sounds vaguely like the cooling system/motors of the telescope); the coquis are my own recording. I don't remember if you can hear those bloody stupid little frogs from the telescope platform or not, but my abiding memory of Puerto Rico is that the basic soundtrack is coquis wherever and whenever you are.

So grab a headset or Google Carboard and enjoy. And if you don't have a headset, grab some 3D glasses and watch it on your PC, using your mouse to look around. And if you don't have any 3D glasses, just watch it in regular 2D 360 mode.

Final remarks : this isn't done. Eventually I want to extend the tour to the lower section of the azimuth arm, but this is quite a complicated place so it will take more time. I'll also probably try and fix some of the more serious known errors. Most irritatingly, I can't seem to get a codec that gives good quality on this, so I suspect you're losing quite a lot of detail. Still tests seemed to give rather better results than the animation, so hopefully the next version will look shinier. I'd like to render in 4k; this one took 80 hours and over 30 GB of rendered files though, so that requires a bit of logistical planning.

Final final remarks : a lot of work has gone into this - tens of hours, if not more - and it's already been used for one commercial product. So unless you're a) an Arecibo Observatory employee or b) someone I've known for many years and already trust, then no, you cannot "just have" a copy of the model. Please stop asking, it's rude. To end on a happier note though, if you ask Bathsheba Grossman very nicely, it's possible you can buy one of the glass cubes. They're very sparkly and make a nice conversation piece. You can't walk around inside them, but at least you don't need a headset.