Articles

Photographing a Black Hole


So tonight, we travel into space to
realms that are replete with mystery. Black holes. They are areas of space that
exert a gravitational pull so intense that nothing, not even light,
can escape from being swallowed. They distort time and space around them. And they come in different sizes,
teeny weeny, middling, and massive, super massive. They’ve been likened to
giant garbage disposals and long, dark tunnels to nowhere. And there are curious terms
associated with them. Spaghettification, singularity, dumb holes, and event horizon. I think all of our imaginations
are spurred by black holes. Einstein predicted their existence, but
we’ve never been able to actually see one. Our special guests, however, bring hope because they
just might be able to show the world what a black
hole really looks like. It’s a privilege for us to present
Shep Doeleman and Katie Bouman together. And here at the Museum of Science. So, please join me in
giving them a warm welcome. First up, Shep Doeleman. [APPLAUSE]
>>Okay, it’s a real pleasure to be here.>>Louder. It’s not on.
It’s on.>>It’s a pleasure to be here. When I told my family that I was giving
a talk here, at the Museum of Science, and I asked them if they wanted to go,
my son said, of course, because for him this is ground zero. This is where all science began. This is where the T Rex is, right? I’m a West Coast boy, so
it’s not ground zero for me. But I know people with children,
this is where all the kids come to get their first real taste
of science in the area. So it’s a really special place. So it’s wonderful for us to be here today, to talk to you about a project
that we are very passionate about. And as the introduction said, we are trying to do something
that has never been done before. We are trying to see the unseeable. Simply put, we are trying to create a new instrument
that can see what a black hole looks like. I think we all know what black
holes are in some sense, right? They are gravity run amuck. When matter gets so dense, and there is so much of it,
that the force of gravity crushes it. There’s no stiffness. There’s no characteristic of the matter
itself that can resist gravity. Everything gets compressed into a point. And there’s a boundary around that
point called the event horizon, where the force of gravity is so
strong, that even light can’t escape. So it’s the ultimate cloaking device,
it struggles not to be seen. And the Event Horizon Telescope Project
that we are gonna talk about today, is trying to do just that. To see the shadow of this black hole. And we hope to see something a little bit
like this picture here, which the swirling gas around and this last orbit at which
lights can circularize around there. And if we could do this, if we could put
high resolution binoculars on and see into the heart of the black hole, that is
where we could test Einstein’s theories. That’s where Einstein’s concept
of gravity might break down. We know it works on solar system scales,
but does it work at the boundary
of a black hole. That’s what we wanna find out. So to set things in context, I wanna show you one of my
favorite galaxies, all right? This is Centaurus A, it’s one of
the nearest galaxies in our area. And all the light you see
there can be well understood, most of this is all star light. And there’s some dust lanes going up and
down here. But when you look at it in the radio,
When you look at radio waves, you see something remarkably different. You see these bright jets of
high speed material streaming from the center of the galaxy. This should terrify you, right? These are blowtorches that go tens of thousands of light years
from the center of the galaxy. The amount of energy in these,
like blow torches, these relativistic jets of material,
is equivalent to one supernova. One star exploding every single
year at the center of this galaxy. So it’s extraordinarily powerful. And the only thing that we know of
that can power something like this, is a spinning, super massive black hole. A black hole that weighs millions or
billions of times what our sun does. Acreting matter and
efficiently turning all that energy of the onrushing material
into radiant light. And that’s what you’re seeing here. So we know that there’s something really
terrifying in the center of these galaxies, and
black holes are really the only answer. So let’s talk for a minute about
how black holes got started. And you can’t really talk about
that without discussing Einstein. When Einstein came on the scene,
there were mysteries. Like the perihelion of Mercury. For some reason, the orbit of Mercury
doesn’t come back and close on itself. It processes. And nobody could figure out why that was. People had exotic constructs and
exotic explanations for that. And Einstein told us that gravity should be understood as
a deformation in space time. Instead of immediate action
instantaneously between pieces of matter. He said matter deforms space time and other bits of matter orbit
in that warped space time. Kind of a geometrical construct
of how to understand gravity. And when he did this, he immediately explained this
perihelion of Mercury problem. It was pretty amazing, right? So we have this new concept
of how gravity works. And just the next year, this man, Karl Schwarzschild did something
with the Einstein’s equations. Only one year later he posited,
what if all the matter were in a point? How would you solve Einstein’s equations? And he saw right away that there was this
boundary called, the Schwarzchild radius, which we now call, the event horizon,
where the force of gravity is so strong that even light can’t escape. So even one year later, that was 1915
was Einstein, 1916 is Schwarzchild. We already know that there’s a one way
door out of our causal experience. And that is also terrifying. Black holes are the only thing ,the
only knot that you can tie and then not untie, right? So one way kind of formation of something. The other thing that is
interesting about Schwarzchild, is that he did this in
the front in World War I. I don’t know about you, but I can’t even operate without
a cup of coffee in the morning. Just the thought of sitting down, working
out Einsteins equations while shells are bursting overhead,
it just boggles my mind. These guys were made
of really stern stuff. But, we really didn’t know that
Einstein was on to something, until, Sir Arthur Eddington created
this new kind of experiment. He said, well,
during a total solar eclipse, the light bent by the gravity of the sun,
by this gravitational well of the sun, should make stars appear to be offset
from where they really are, okay? So if we’re here on the Earth,
we would see the star here. But it’s actually over here. And he led an expedition off
the coast of Africa to Suriname. And he saw, indeed,
the offset positions of these stars during a total solar eclipse, were exactly what
Einstein had predicted and not Newton. And overnight,
Einstein became a household name. And Einstein’s theory of gravity
is what we use even to this day. So Einstein’s theory is working well,
but then the question is, So how do you make a blackhole? If you had a laboratory and
you wanted to make a blackhole, you’d have to crush things. This table here, it’s not gonna surprise,
hopefully many of you, to know that most of this is empty space. It’s hard, [SOUND] but
the distance between the atoms is vast. It turns out that the sun,
if you look at the density of the sun, it’s about the density of water. There’s nothing really
crazy about that density. It’s about one gram per cubic centimeter. It wasn’t until people like Chandrasekhar
looked at white dwarves, that we found that nature can squeeze all the space out
of an atom until only the fundamental particles are left, and there’s a star,
Sirius D, which orbits around Sirius A, the dog star, which is the mass of
our sun, but the size of the Earth. That’s a white dwarf, and
it’s held up by quantum principles that prevent electrons from being
too close to each other. This is intensely dense stuff. It’s 100 million grams
per cubic centimeter. So, right away we understood that some of
the things we’re used to thinking about on Earth don’t really hold in space. That there are crazy,
extreme objects out in space, but even this is not a blackhole. So how do we understand some
of the extremes of matter? Well, our sun,
not anytime soon by the way, we have a couple of billion years left,
is gonna turn into a white dwarf. Above the mass of our sun,
you can get a neutron star. So even all the electrons, and protons inside the white dwarf
can fuse together to be neutrons. You can get even denser, and that is the
situation where you have a star that is the mass of our center, but the size
of New York, or the size of Boston. One teaspoon of it weighs
one trillion tons. Hugely, hugely massive. But above that, people like Oppenheimer
and Snyder in the 1930s worked out that even neutron star material,
when there’s enough material around it, it cannot withstand the force of gravity,
and those objects become blackholes. So we have now a way for
stars to become blackholes, and those are the small blackholes that
Lisa was telling us about earlier. But there are also super
massive blackholes. They’re the kind of blackholes that exist
in the center of that galaxy that I was showing you before. This is a different galaxy. This is M87, and Invirgo A, and you see
a jet like you saw before coming from the nucleus of the galaxy, and when the
Hubble Space telescope looks at the gas, right around the center of that galaxy,
you see that there’s a red shifted part coming or going away from us, and
a blue shifted part coming towards us. And the only thing that can
cause that orbital motion, so close to the center of the galaxy, has to be something that weighs about
7 billion times what our sun does. So again, very strong evidence that there
are super massive blackholes in the center of galaxies, and we understand this
now because galaxies can merge over the lifetime of the universe, and medium
size blackholes can turn into larger blackholes, and so on, and so on until
you get billion solar mass blackholes. So why do blackholes glow? We’ve talked about they’re the ultimate
sinks, they’re garbage disposals. The reason they glow, the reason in a
paradox of their own gravity that they’re some of the brightest things in the sky
is that they really efficiently turned gravitational potential
energy into heat and light. They’re really small and
they have huge attractive capabilities. So all of the gas and dust around them
are trying madly to get into a tiny, tiny volume. Just as when you would rub your
hands together to get warm, all the friction heats up that gas to
hundreds of billions of degrees, so they shine very, very brightly, and this
conversion can be 40% efficient, that’s compared to a nuclear bomb or a nuclear
fusion, which is only about 1% efficient. So there are terrifyingly
efficient engines for turning gravitational energy into
radiant light, so that’s why they glow. This is what we think the emission
around the blackhole might look like. I’m gonna get that going for you. So this is a simulation done by
someone who works in our group. Let’s see if we can get it going. Maybe it’s not gonna animate for us. There it goes. So this shows the swirl of gas
surrounding the blackhole, and we wind up seeing here,
I want you to pay close attention to this, is this ring of light
that seems stationary. It turns out that that marks the last
orbit at which light can move around the blackhole. Even light is bent by gravity, and even if
you’re travelling at the speed of light, you’re forced onto this circular
orbit around the blackhole, and the reason you see the shadow is
because much of the light that’s emitted by the gas surrounding the
blackhole, is lensed onto that last orbit. So you wind up seeing a ring
with a relatively dim interior, and that’s what we call the shadow, and Einstein’s equations predict the precise
shape and size of that shadow. These are equations that
he wrote 100 years ago, and this is what
the Event Horizon Telescope is after. If we can measure this ring, if we
can detect that size, and that shape, we could test that Einstein’s theories
hold at the boundary of the black hole, and in the case of a blackhole
I’m about to show you, we would know that there were 4,000,000
suns of mass inside of that ring. So the best candidate we have for a super massive blackhole that’s
near us that we could resolve and make a picture of, is Sagittarius A star
in the center of our own Milky Way Galaxy. It’s about 25,000 light years away. This shows streamers of hot gas flowing
into the center of our own galaxy and this little white spot here is the point that
marks a 4 million solar mass blackhole. We know that because there’s been
some extraordinary work done by astronomers working in the infrared. They’ve looked at stars that emit in
the infrared, and you can see that there’s a unseen mass here just tossing
stars around like they were planets. This orbit in particular, SO2 is an 11 year orbit that has now been
seen to go all the way around and close. What can cause a star to
orbit it like a planet does? It turns out that the mass of this unseen
object there is 4 million solar masses, and this is extraordinary stuff. This knocked my socks off,
and many other people’s too. They were all sockless. When we saw this this was a real turning
point, because this tightened the noose incredibly, and yet
we haven’t seen a blackhole, all right? They’re like dinosaurs.
We know they exist, we think they exist, but wouldn’t it be great to
see a picture of a dinosaur? Imagine you could see something like that. That’s what we’re after. So, as they say in show business, so
you wanna photograph a blackhole shadow? Well you need the highest
magnifying power ever assembled, because the smallest size we can observe
on the sky is equal to the wavelength of light you’re observing divided
by the size of your telescope. Standard, kind of, simplest equation you
can write for this kind of stuff, and the shadow size is 50 micro arc seconds. Let me decipher that, that’s like
seeing an orange on the moon, or, as Katie reminded me
earlier this afternoon, it’s like being able to hold a single
atom at the end of your arm, and see it. So that’s the kind of angular resolution, that’s the magnifying
power we need to conjure. But we also need to see
Sagittarius A stars. We have to see through
the Earth’s atmosphere, the interstellar gas between us and the
galactic center, and we also have to see through that 100 billion degree gas
that’s surrounding the blackhole. And to do that, the only game in town
really are millimeter wavelengths, about a couple hundred gigahertz, for
those of you who think in frequency. So, if we put into this equation here,
the wavelength of one millimeter And the smaller size being 50 micron seconds,
which is the size of that ring. We come up with the size of
the telescope being 10,000 kilometers. We have to build a 10,000
kilometer wide telescope. And as we’re scientists and we kind of
MacGyver things, that’s what we do for a living, we just say, we’re gonna do it. And we strap on our utility belt,
and we just get to work. And that’s what the Event Horizon Tlescope
really is all about. You can’t build a single telescope
that’s the size of the Earth. And the way these single telescopes
work is like a big radio dish, right? But it operates the same
way an optical does, is that light coming from the cosmos hits
this perfectly defined parabolic surface. And the surface is tuned so that all
the light gets reflected to this receiver. That’s where you put your camera, and
that’s how a normal telescope works. But we need to make something
that is 10,000 kilometers across. And the way we do that is with something
called very long baseline interferometry, or aka the secret sauce of
the human horizon telescope. What we do is we take radio telescopes
that exist already at different parts of the Earth, and
we look at the same object on the sky. And we record the light, we freeze the
light at different points on the Earth. And we use an atomic clock to time tag,
so we know exactly when the light rays from this black hole or this object
hit that particular point on the globe. We record it on, it says magnetic tape,
we used to do magnetic tape. We record on hard disks now, and
we bring it together and compare it. This comparison is exactly
the same operation as light bouncing off a surface and
combining into focus. But we do it in a supercomputer, what
a mirror does just by its geometry, okay? So the name of the game is to
record as much light as you can, time tag it perfectly and compare it. And then you wind up getting
a data set that’s equivalent to having a telescope as large as
the distance between these two dishes. And what our group has done is work on the
technology breakthroughs that have allowed this to happen. So instead of reel to reel tape recorders, we’ve developed these
banks of hard disk drives. And we record many times the amount of
data in the Library of Congress every night that we observe. And instead of a bunch
of analog electronics, we use these single chips here that do
all of our signal processing for us. That make the data ready and
turn it to ones and zeroes that we can write onto these disks. And this has catapulted things. We’ve used Moore’s Law, so that we have
moved things almost a factor of 100, in terms of the data that we’re
recording at all of these sites. The first thing we did
was we wanted to ask, can we see something the size
of the event horizon? And so the first experiment that we did,
we linked telescopes in Hawaii, Arizona, and California, and
we looked at the center of our galaxy. And what we expected to see
on the short base line here, on the short distance
between these two telescopes. Remember, it’s lambda over
D is the magnifying power. D is fairly short here,
it’s only 908 kilometers. That means you’re looking at
a big section of the sky. So if the emission around
the black hole was finite in size, you would still get all of that emission,
okay? This baseline would be sensitive
to all of that emission. But on these long baselines,
the magnifying power is so strong that you’re looking at
a much smaller part of the sky. So if the emission region was this big,
you are only looking at part of it. You’d only expect to get
a fraction of the power. And that’s exactly what we saw. This is the only graph, I think,
that I’m gonna show in this talk. But this is the power here,
and it’s the baseline length. It’s the distance between the telescopes. On the California to Arizona baseline, we saw all of the power
expected from this black hole. But on these long baselines, and
this curve shows what we expected to see, you see much less power. And that’s indicative of us resolving or seeing only a fraction of
the energy from the black hole. We sized this black hole,
we measured its extent and it was about four
Schwarzschild radii across. About exactly what we think
the shadow should be. This is a kind of a-ha moment that I wish
everybody should experience in their life, no matter what field they’re in. Cuz when we saw this graph, we put this
together, we knew what we had to do for the next 20 years of our life, right? We had to continue on and go from a few
telescopes to a much bigger array, that now involves telescopes all over
the world including the South Pole. And when you get this many telescopes
together, with this kind of technique, you can move from simple size measurements
to making a true image of the black hole. And that’s what Katie is
gonna talk about right now.>>[APPLAUSE]>>All right, thanks, Shep. As Shep mentioned, there’s many parts that go into
getting this picture of a black hole. But kind of the last stage is taking the
data and trying to make an image from it. So I’m gonna try to give you
intuition of how we do that. So remember that Shep said that if we
wanna take a picture of the black hole in the center of our galaxy,
given what we know, we would need to build this impossibly
large, Earth-sized telescope. So let’s for a second just imagine we
could build a telescope that is the size of the entire Earth. A radio telescope, as Shep said, acts
a lot like a mirror, or it is a mirror. And so light traveling from the black
hole travels to Earth for 26,000 years. And eventually it will get to Earth. And if we had a telescope dish like this,
the size of the Earth, it would bounce off the dish and
go to this focal point. Now, the key to making it possible for
an Earth-sized telescope to make a picture is that the light from the black
hole would bounce off many locations and combine together at this single location. And if we could do this,
we could just start to make out the ring of light that is indicative
of the event horizon of the black hole. So although this picture wouldn’t look
like all the computer graphic renderings that we’ve been able to do, we’d be
able to safely get our first glimpse of the immediate environment
around a black hole. So clearly we can’t build a telescope
that is this large, but instead, let’s think of a different
way that we could do this. One way is let’s try to put
mirrors all over the entire Earth, essentially turning the Earth
into a giant spinning disco ball. So in this case, the light would
also travel from the black hole and reach our disco ball Earth. And maybe at each little mirror
we could collect the light, recording what we’re seeing. And then instead of it all bouncing
back to one place where it’s combined, instead we combine it with a computer,
all of the recorded light. And doing this, we’re able to get an image
that looks just like if we had a giant dish the size of the Earth. Okay, so by turning the earth into
this giant spinning disco ball, we can get measurements that can give
us a picture that is just as good as the Earth-sized dish. But now let’s imagine if we
removed a lot of these mirrors, so only a few of them remain. Now there’s a lot of holes and we’re only
collecting light at a few locations. So these remaining mirrors represent
the locations where the event horizon telescope has telescopes. And using these measurements to
make a picture is really hard, cuz there’s such few measurements. But luckily, as the Earth rotates,
we get to see other new measurements. In other words, as that disco ball
is spinning, the little mirrors change locations, and we get to
see different parts of the scene. So using the sparse data that we collect,
the imaging algorithms that we develop fill in
the missing gaps of the disco ball. To reconstruct the underlying
black hole image. But you might look at this and be,
well, there’s a lot of missing data. How are we filling that in and making sure that we believe
the image that we that we get. And so to give you kind of
an idea of how we do that, I wanna give this analogy where you can
kind of think of the measurements that we see from the telescopes
a bit like notes in a song. So each measurement is like one tone or one note [SOUND], and
different measurement that we take from the telescopes is
a different note [SOUND]. So the image of the black hole that we
see with the Event Horizon Telescope is a little bit like listening to a song
that’s being played on a piano with a lot of broken keys. So although if you were
to hear a song like this, it’d be kinda hard to make out what
the song is, but a lot of times, you’d still be able to get
the general gist of it. And so
just to make this a little bit more clear, I wanna show you an example of
as you add notes to a song, how you’re able to kind of fill
in the missing data yourself. So what I’m gonna do is
I’m gonna play a song and I’m going to increase the number of
notes that you hear, just like as we’re increasing the number of telescopes
in the Event Horizon Telescope array. And hopefully, eventually, you’ll start
to kind of make out what the song is, or you’ll at least get the idea of the beat
of the song, hopefully [LAUGH]. So I’ll also light up the keys
that I’ll be playing. So at first,
there’s only gonna be one note playing. Okay, so ready? [MUSIC] [LAUGH] [MUSIC] Okay, so by close to the end [LAUGH],
hopefully, you were able to kinda
get what the song was. And if you don’t know the song Ice Ice
Baby, then at least maybe you could kinda get the beat or kind of fill in
the missing information that we had. And it’s kind of really amazing that we’re
able to do that because there were a lot of notes missing there. But yet our brain, despite that, our brain
is able to fill in the missing information and we can kind of figure
out what the song is. So no one is telling you that
the notes that we weren’t playing. Maybe someone could have just been banging
on the notes that we weren’t playing, and it would be a totally different song. But that’s not a really reasonable way to
put in the notes, to put in the missing information, and so you can kind of
figure out what the best song is. And what your brain is doing here is very
much like what the imagining algorithms that we developed for
the Event Horizon Telescope are doing. There are a lot of possible images that can perfectly explain the telescope
measurements that we make, but some of them are just more
natural than other ones. But there is one thing I kind of
want to bring your attention to and that’s that there is always some ambiguity
to the measurements we’re making. So even at the end when there
was a number of notes playing, maybe a number of you
knew what the song was. It didn’t have to be Vanilla Ice’s Ice Ice
Baby, it could have been some other random combination of notes that were missing
in just a completely different song. And the less notes that we see,
the more ambiguity there is. So maybe at this point [MUSIC], some of you even had confused it for
Queen’s Under Pressure [LAUGH]. So [LAUGH] if these were our measurements,
we’d be in a bit of trouble. There’s two songs that fit
the measurements fairly well and so we can’t make a good
judgment on what it is. But to be fair, I chose this song for the
reason that I wanted to kinda demonstrate, that I wanted to demonstrate
that there is this ambiguity. And this ambiguity also [MUSIC], I’m sorry [LAUGH]. That’s Under Pressure,
if you don’t know [LAUGH]. So this ambiguity exists
in the measurements that we make with the Event Horizon Telescope. So similarly for
the Event Horizon Telescope, the data we take only tells
us a part of the story. There’s still an infinite number
of images that perfectly explain the telescope measurements that we make,
but not all images are created equal. And some of those images look more like
what we kinda think of as images than other ones. And so what we do is we
rank images based upon how likely we think an image is and then
we choose the one that is most likely. So I’ll come back to what this
means a little bit later, but by using kind of this method, we can then make reconstructions of
the data from the Event Horizon Telescope that also match the data,
that also look like reasonable images. So here,
I just show you a simulation done using simulations of telescopes from the Event
Horizon Telescope array when we pretend to point our telescopes towards the black
hole in the center of the galaxy. And although this is just a simulation,
reconstructions such as this give us hope that we’ll soon be able to reliably
take the first picture of a black hole and also from it,
extract the size of that ring. But the question remains of
how we actually define what a reasonable image is. What is a good black hole image? So let’s take a step back. Instead, let’s first say, well, how do we
define what a likely Facebook image is? So many of us have experience with
Facebook, we’ve seen what people post. And so we can try to look at an image and say, well,
how likely would you see it on Facebook? So it’s pretty unlikely someone would
post this noise image on the left and it’s pretty likely that someone would
post a selfie like this one on the right. The image in the center is blurry, so even
though it’s more likely than the noise image, it’s probably less
likely than a selfie. But when it comes to images from the black
hole, we’re posed with a real conundrum. We’ve never seen a black hole before, and
in that case, what is a likely black hole image and what should we assume
about the structures of black holes? We could try to define
what a likely image is by looking at the simulations we’ve done
like the one Cheb showed earlier, but if we did that, that could cause
some really serious problems. What would happen if
Einstein’s theory didn’t hold? We’d still wanna reconstruct an accurate
picture of what was going on. And if we bake those pictures and
theories, the theory too much into our algorithms, we’ll just end
up seeing what we expect to see. In other words,
we wanna leave the option open for there being a giant elephant
at the center of our galaxy. So what we do is develop algorithms that
try to define what is a reasonable or likely image And these need to define it in a way such
that it’s not so restrictive that it only tries to find something that we expect
to see, like the black hole shadow. But also are able to get images kind
of like the elephant in the center of the galaxy. But there are many ways that we
can define what is a likely image. And so we develop multiple
algorithms that kind of try out all of these different
imaging assumptions. So for instance here I might
show you three methods. Each one kind of defines what is
a good image in different ways. So method one might say okay,
I like blurry, kind of fluffy things, those seem
like things that I expect to see. Method two might say,
I like images that I’ve kind of, that have features of things I’ve seen
before in this world and in space. And method three might say, I want
an image that has really high contrast. So all these methods kind of have
different assumptions on what an image looks like, but
if given the same data, they all produce a very similar looking
image with the same kind of structure, then we can start to become confident that
the image assumptions we’re making aren’t biasing the picture of
the black hole that much. However, in the case when one or
more of the algorithms produces an image that has a different structure, then that
alerts us that there might be a problem. In this case, often the data
is not enough to constrain us. And so the image assumptions
start to take over and we can’t really trust what we’re seeing. But it’s incredibly important
to us that we’re accurately representing the structure of the black
hole that we’re trying to image and not just trying to make pretty pictures. So to make sure that these methods work,
and to make sure of the way that we can
verify image works, we’ve been testing ourselves using something called
the Event Horizon Imaging Challenges. And the imaging challenges give us a way
to test out the entire imaging pipeline before we have to do it for
real just in a couple of months. So in the challenge, we have a team
of people who selects an image and from it generates data that is
very much like what we would expect to see from the event
horizon telescope. It has all the different kind of
noise properties and everything. And then they make this
data publicly available so that people who have
developed methods to be able to take this data and make an image,
are able to make an image from it. And these are many different methods
with many different assumptions, but all of these people do not have any
access to the ground truth image. So, they’re just making their best
guess at what they think the image is from the data that they see. And then we take this data, these images,
and we send it to a panel of experts, some astrophysics experts, and
we ask them, what do you think is real? And this kind of simulates a little bit of
the process of what we’re actually going to have to do with the real
data that we’ve collected, but differently from the real data we’ve
collected, we also have the true image. So we also can compare how well our
images match the actual true structure. So I wanna show you a few examples of
some results from the Imaging Challenge. So here you can take a look
at one such example. Here I show five images that were
submitted from five different methods. And, the telescopes do not see color,
so we only get grayscale images. But I’m gonna show it in
this other color map just so you can see a little bit
more of the structure. So, before I show you the true image, I just want you to kind of
think about a few questions. For instance,
what do you think is real here, what do you think is a real structure? Do you think that the images
are consistent with each other? Do you have any guess to
what the actual image is? So if I look at it, I actually think
that the images look fairly consistent. There is a kind of ring structure,
and it’s kind of bright on one side. But one thing that I’m a little
bit more concerned about and not really sure of is that four of these
methods, it’s maybe a little hard to see, but four of them kind of have
a tail coming off the end. In one of them, it’s missing. So if I see something like that
where not all images are consistent, not all showing the same feature,
I’m a lot more skeptical of it. Okay, so let’s see what the real image is. Okay so actually the algorithm seemed
to produce a pretty good image. They all kind of get that
ring that we’re seeing. And even though all the images
look quite different, for instance this one kind
of has very sharp edges. Well, okay,
the one on the left has very sharp images. And the other ones
are a little bit more blurry. Then they still all kinda get
a very similar structure. Okay, so how about one more
example from the challenge? Okay, so here is from another
set of data that we had. Where the people who were generating
the data used a different image. Okay, so I’ll once again show
it in this other color map. And again ask yourself some questions,
what do you think this is? Do you have any idea? Do these look similar? Do you believe them? Okay, so, you guys ready? Okay, well. [LAUGH] This time,
the image was actually Frosty the Snowman. So, although, I don’t think anyone
expects us to have a huge giant snowman at the center of our galaxy, I
think this was a really fruitful exercise. Because although this is not expected,
we want to be able to have our algorithms be flexible enough to
actually see something that is unexpected, like the elephant in the center
of our galaxy or the snowman. And I am really happy that the algorithms,
even though they are not usually tested on images like this, still are able to get
the general structure of the snowman. And even in some cases
get the little arms. [LAUGH] So yeah, as Shep mentioned, earlier this month actually, we took the
first measurements from the event horizon telescope that we’ll actually possibly
be able to make an image from. So I’m really excited for
us to be able to get this data very soon, get our hands dirty and
start making these images. And hopefully as our imaging
algorithms progress, we may even be able to start
making videos from them. So these imaging algorithms will
hopefully make it possible for us to test Einstein’s famous equations for
which scientists rely on a daily basis. And hopefully learn more about
the dynamics around a black hole. All right,
I think Shep wants to say a little bit of something about, yeah.>>[APPLAUSE]
>>[LAUGH]>>Well, while we’re both up here, it’s really important to
understand that these are now two pieces of the whole project. I’ve walked you through black holes and some of the fundamental principles
of the event horizon telescope, how two telescopes can act as a telescope
the size of the distance between them. And then Katie showed us very
intuitively how you take many, many telescopes and synthesize an image. And it’s all kind of academic isn’t it? Can we do it, can we not do it? And again I wanna show
you that we have many, many telescopes now in the event horizon. Project, and as Katie said, last month, after a ten year preparation period,
we took our first data that involved really this,
where are we now, there it is, which involved the center array in Chile,
is that cluster of telescopes in there, and that increased
our sensitivity by a factor of ten. So imagine, all of a sudden,
you could see things times ten. So, the earlier work that we’ve
done is now greatly amplified, and we really think that we have a very
good shot at making the first image of the Event Horizon, and
that is quite extraordinary. It’s a really an interesting place and
time to be in the project, and I wanted to say Terry,
that the whole team is really excited. So we have people here at the South Pole
that are getting the telescope ready there. We had people in Mexico
working on atomic clocks. These atomic clocks only lose one second
every 100 million years, so they’re very precision instruments, and they have to be
at every single one of these sites, and your heart hasn’t skipped
a beat until you’ve winched one of these atomic clocks up a helical
staircase at 15,000 feet. That really gets you going, and
these are people also in Mexico, these are people down in Chile,
these are our colleagues in Spain, these are some people up at the top
of the highest mountain in the chain. So it’s really hats off to these people,
and to the whole team, and we’re just two representatives of the team, and
we think we’ve very, very close to it. So stay tuned, and we hope to have something to show you in the near future.>>[APPLAUSE]
>>Again, it is somewhat academic, so I think we’re
gonna sit and chat for a minute, Katie and I, just to give you a little sense of
the human element of this, and the reason is that Katie was at a remote site during
part of the experiment we ran in early April, and I was at a central command
facility here, a bunker at Harvard, right? Harvard has bunkers, and they are not
underground, on the second floor, and, as we had different experiences, and
we’re just now actually, this is one of the first times we’re seeing each
other after all these activity. So I want to start off by asking you
Katie, to explain to me and to us, how did you get into this field,
because, I know how I got into it, but why don’t you say a few
words about how you got into it?>>Sure.
>>Or do you want me to go first?>>No, that’s fine. Yeah, so I actually come at this,
I’m not an astronomer. I actually had the great
opportunity to work on this, because I got in touch through my
advisor who, we do computer vision. So I work at MIT. I’m a PhD student who studies computer
vision, so I work with pixels, and trying to understand what’s going on in
images, and I think, I had the opportunity to kind of travel up the Haystack
Observatory I don’t know, three years ago? [LAUGH] Or four, long time ago,
[LAUGH] I guess now.>>I should say Haystack Observatory
is where that supercomputer is, that merges all the data from all
the sites of the Event Horizon Telescopes. So that’s kind of a central node for
us and we make pilgrimages there.>>[LAUGH] Yeah.>>Leave burnt offerings. Our and chocolate is always useful there.>>At the time I knew
nothing about blackholes, but I heard that there’s a project that
maybe could use an imaging person, and so obviously,
I jumped on the opportunity. In computer science, we don’t often get
the opportunity to work with blackholes. So I went up there and it was just and others there who described
this really cool project, and I left understanding about 10%
of what they said, if that, [LAUGH] but knowing that I really
wanted to work on this project, and at the time I didn’t
really know too much about it. But I kind of started reading about it and
everything, and I started realizing that actually, the
problems even though it all centers around everything about blackholes and stuff,
but really it’s not that different from a lot of the problems the we deal with in
computer vision and in image processing. We have this data,
this really, really noisy data, sparse data, and we wanna try to
get out something from it, and these are problems that I kind of work
on all the time in our field and, for instance, medical imaging. When you go and
get an MRI of your brain, or something to figure out if
something is going wrong, it uses a lot of the same ideas that we’re
using to take an image of a blackhole. So it’s really not that different,
but I’m so glad that I was able to work my way in,
[LAUGH] and try to have a small
contribution to the project.>>And don’t let her fool you. Katie is contributing
mightily to the project. She runs the entire
Event Horizon Telescope imaging challenge, which is really an incredible feat. And so, for my part, I got into this,
I wasn’t a boy astronomer. Some people grind their own lenses,
and they make their own telescopes.>>[LAUGH]
>>They want jet packs, and I came to this late in the game, and
when I graduated from undergraduate I saw this flier and
it said do you want to go to Antarctica? Is there any other response than yes? And so I sent this application in,
and I was interviewed, and they said you’re hired. And I said, great, what do I get to do? We’re gonna send you there for
a year, and you can’t leave, right? But that was probably one of the best
things that ever happened to me professionally. I was quite young, and I spent a year on the coast of Antarctica,
and took care of a lot of experiments, cuz that’s where the magnetic field lines come
down and they pierce the Earth’s skin, and so there is a lot of interesting
phenomenon to study there. And I was put in charge of
all these experiments, and there’s no Radio Shack down the road. There’s no drones from
Amazon coming in there. If something breaks you have to fix it,
so you get really good with duct tape. You get really good with binder twine,
and pliers, and there was a real joy in doing science in difficult circumstances,
and that is what does it for me. I’m not a computer vision person, but
I absolutely am in love with going through the telescope 15,000 feet and
doing things that nobody’s done before. So that’s my tie in to this,
and then the other thing that we were thinking about
was a different role. So, like I said, I was in a central
command facility here, and Katie, you were at
the Large Millimeter Telescope. So what was that like? I’ve been to the LMT, but what was it like
this time to make the observations there?>>Yeah, well-
>>And that’s this telescope, the center one there
with the atomic clock.>>So we had a big team of people
at the Large Millimeter Telescope. So, as Shep said, there was groups of
people that travelled to every site. So we had people at the South Pole,
people in Hawaii, Arizona, and I was lucky enough
to get to go to Mexico. And so a large millimeter telescope
is situated at about 15,000 feet. So we work up there all day. It’s a little bit draining cuz you
have not as much oxygen up there. But it’s a huge amount of fun. So as I said, I don’t really, I originally
didn’t have much experience with telescopes but I did kind of was
interested in the imaging side of things. But one thing that I think
is really important for the imaging is trying to understand
exactly what the data is, where it comes from,
what kind of noise enters the process and I think all these kind of things and as I
work more with the team and I’ve gotten to spend more time at the telescope there I
get to learn a lot of this kind of stuff that I don’t get just by looking at the
theory of it here [LAUGH] in Cambridge. So it’s really cool. So my role I guess at the telescope was,
I was helping mostly with taking the light I guess, and
digitizing it on to those hard drives. So that we could then fly it
back here to kind of treat, to process it like we kind of saw
the light all at the same time. From all the different locations. So it was pretty cool, we had to, I guess a lot of times we
have to do the observations at night. So we had to switch our schedule,
so we’re sleeping during the day. Unfortunately, the gas truck [LAUGH] and
the roosters crowing and stuff kind of keep you awake but then we go up at night, and
we’d take the measurements then. So how was the central command? I know that i used you guys a lot
[LAUGH] when I was at the [CROSSTALK]>>It was a little different the central command form because there
are problems in the field, right? You get these emergency calls, right. So all of this in
a recorder isn’t working. And these recorders are vital, right? We’re recording 32 gigabits per
second at every one of these sites. So we record almost a petabyte of data. 1,000 terabytes of data at
each one of these telescopes. No internet could ever send that back. It would take about 40 days,
2 months just to send it back constantly. From the South Pole we only get 50
gigabytes per day allotment of data. It would take us 24 years to
get that data back, right? So the fastest thing you can do is
load up the 747 with disk drives. No internet beats that, alright. So we use FedEx, and FedEx it turns out, the best thing
that ever happened to Radio Astronomy.>>[LAUGH]
>>So one thing that we do, at the central command,
we’re the clearing house. So we are 24/7 open. We have lines of communication open
to all the sites, LMT included. And whenever somebody had a problem, we would route the problem to
the local expert who was at sea level, who was rested,
who had just eaten a nice steak sub. You know, who was just like,
happy to sit in front of a computer and help somebody remotely. And that’s really important because at the sites you get what you
call summit moments, right? I don’t know if you guys know what that
means, but when you’re at 14,000 feet, there’s not enough oxygen and
you start getting loopy. And I really do mean loopy. And I remember standing on a ladder
once trying to screw something in. And people are down below me
saying would you please hurry up. And I’m saying I’m really
working on it here. And it was only after about a minute
that I realized I was unscrewing this thing, right? I was just wrong direction. That just happens up there. So you need somebody at sea level and
that was what we were doing at the central command, and the big thing for
us making the go, no-go decision. Cuz when you think about it, you need the weather to be good
everywhere across the array. So normally when you do astronomy it’s
enough to have one telescope with good weather. We need eight telescopes
with good weather. And sometimes you’re left watching a cloud
or a big massive water vapor drift towards your telescope and you’re just,
is it gonna pass, you know? Is it gonna hang out over my telescope? Is it gonna form ice on the dish and
make us shut down? And you make a call, go or no go, and those are the most agonizing things
we do in the entire project, frankly. You know if you say go and
then the weather closes in, it’s horrible. You’ve wasted a night. And the most precious
thing we have is time. Because these telescopes we use are over
subscribed to particularly by factor of tender one. And you have to compete
like mad to get that time. So if you lose, if you waste
a whole night that’s really gold. And you feel as bad frankly, when you say
don’t go and the weather’s beautiful. Right, cuz you’ve just
wasted some time there. So that was the general difference
between being at a remote site and I think being in the central bunker.>>I wouldn’t say you guys were so
well rested though. We woke you up plenty of times at like
3:00 AM [LAUGH] asking questions.>>That’s really true, that’s really true. In fact,
once I woke somebody up in Holland, because they were the local expert
on this piece of equipment. And he was not really prepared for it. And he said, well I’ve left my
password at work, you know? And I was like we’ve waited a decade for
this buddy [LAUGH] you get to work, you know?>>[LAUGH].
>>You get your bicycle or whatever. [LAUGH] You bike to your
dutch place of work and you figure out what that password is and
you use it. Speaking of which we’ve talked about our
different ways in which we joined this. And by the way,
if you ask a bunch of kindergarteners what they want to be when they grow up
it’s all like ballerinas and doctors. I guess firemen and stuff. Nobody ever says, I want to image
of black hole mega order, right? So the pads we’ve taken really illustrate
the different ways you wind up kind of at the same place. The last thing I wanted to add is that
there’s been some talk you know recently, politically about difficulty of ideas and
people moving across borders and intellectual freedom. And what I love about this project, is
that it elegantly side steps all of that. All right, it simply and straight
forwardly gets a bunch of scientists together united by a common vision to see
the unseeable, to test Einstein’s theory of the black hole boundary, and it uses
allocated resources from around the globe. In a very interesting innovative way,
to create a new kind of instrument that no one country can own, clearly, but
that all countries can contribute to. And the free flow of scientific
ideas across that make it possible, and it’s really the free flow of people,
too. So we just do our business and we are creating something with this
global community and this global team. Which I think is very special. So it’s a real privilege to
work with the whole team. So did we want to do anything else
before we open up for questions? We are at your disposal.>>Yeah, [LAUGH].
>>[APPLAUSE]>>Alright, so please wait and if you have a question, raise your
hand and one of us will come and bring a microphone over to you. So we’ll have the first
question right over here.>>It sounds like you guys are already
getting started with trying to get a picture of the black hole. When you guys do you guys have
an expectation of when the first time you would actually be able
to definitively see that ring or is that something that
is not planned out yet?>>I would say that we’re
probably a year out. They’re there to give you a real idea. Because we have to bring the data together
and the disk drives are just still arriving at the two supercomputer centers
where they’re gonna be processed. One is in Bonn, Germany, and
one is in Haystack Observatory, about 40 miles north of
where we’re sitting now. And I don’t think I’m breaking
confidentiality to say that we have some very interesting and
good news so far. I mean, it looks like the Event
Horizon Telescope didn’t work, right? This technique is horribly nerve wracking,
right? It’s the ultimate in
delayed gratification. You don’t even know if
your telescope worked until all the data come back, right? And so now we’re just getting the first
inklings that the telescope worked and that it worked spectacularly well. So we have every indication and
we’re optimistic that we’ll be able to use the techniques that Katie was
talking about with this new and very unique set of data. A year.>>So a two part question, one is when
you’re using telescope’s different positioning of it, based on the movement
of the Earth, you’re obviously getting those images, or the data for
those images, at different times. So they’re not all synchronized. So how do you take that into account? Do you just assume a relatively constant
image and try to patch it in that way? The second part of the question is, how
many different frequencies can you use? Not just different colors of light, but across the entire
electromagnetic spectrum? How much more information
do you get by doing that? Or are you not able to do that?>>Do you want, I can, Yeah, and
then maybe Shep can add on at the end. So yeah, you’re exactly right. As the Earth is rotating,
we’re getting measurements. So if there is a lot of motion in In
the black hole, then we’re basically kinda getting data from different snapshots,
different frames of a movie. And so yes, usually in these algorithms we assume
that the black hole is pretty static. And this is a pretty good
assumption in a lot of cases. For instance, we’re not just interested
in the black hole in the center of our galaxy, but also the one
Shep showed in the center of M87. So M87, we think that the black hole
there is very, kind of, static. And it doesn’t change too much
over the course of a night. So this is a reasonable assumption. But for Sagittarius A*,
which is the center of our own galaxy, we actually think it evolves
over the course of minutes. And so we do kind of have to
kind of take that into account. And so that’s why right now we
kind of assume static images. But the next stage is to try to even
improve these imaging algorithms, so maybe we can even make movies from
them and see the dynamics over time. It’s already hard enough to make an image. So you have so much more information
you’re trying to fill in for a movie. But I think that if you model it that way,
we could get some rich information out. And as far as frequencies,
yes I think there are. But maybe Shep will,
maybe you want to talk about that.>>So we’re basically digesting and capturing about 16 GHz of data, okay? I’ll tell what that means in a moment. It’s a slice of frequency,
and we are at 200 GHz. So the fraction of the frequency
spectrum that we’re capturing is really tiny compared to the frequencies. So it’s almost monochromatic. It’s almost as though you’re looking
at one frequency of light, okay? And we can get more sensitivity by
increasing that bandwidth, right? So w can use Moore’s law, the same kind of law that let’s
the computer on your desktop run faster. We use commodity off the shelf pieces of
equipment to build our systems, right? People used to build these by hand,
would hand tool them like leather, right? I mean, they used to make these immensely
complicated analogue systems or hand design them or forge new kinds
of integrated circuit chips to do it. Now we just go to Newegg. And we go to Amazon, and
we order these things off the shelf. And we wire them together. And it’s the sheer raw computing
power of commodity electronics that has let the EHT do what it does now. We maintain the radio waves because
those are the ones that can see all the way through to
the center of the galaxy. You could never do this,
let’s say, in optical light for a number of different reasons. One of which is that you cannot
see the center of the galaxy. So it’s kind of a Goldilocks problem, millimeter wave radiation radio waves
can see deep into the gravity well. And through all the interstellar medium,
all the dreck between us and the galactic center and
through our Earth’s atmosphere. And they just happen to be what we
need to make an Earth sized telescope with the magnifying power that
can see what we are after. Sometimes nature is crazy but
sometimes she’s kind.>>Great, we have the next question
in the back of the house to your left>>Hi, hello, I’m sorry [LAUGH]. I was just wondering, what is the typical
exposure time for an observation run on any of those telescopes to actually
get enough data to reconstruct images?>>Yeah, so this observation run
that we did just a couple weeks ago, we observed over five nights. And each night was,
most nights were 16 hours, so actually, kind of,
as the earth is rotating for 16 hours. And the reason we don’t do
the full 24 hours is because the radial telescopes work
best when it’s dark out. When the temperature is constant, and the dish doesn’t kind of warp
due to the sun being on it. So if the sun is beating down on it,
the dish kind of warps. And it kind of just makes your
picture out of focus and everything. But 16 hours was actually
quite a long time. And so for
us at the large millimeter telescope, because we’re up at such high altitude,
we can’t be there for that long. So we actually would split
our team into two teams. And so I was in the second team,
so I’d go up around midnight and come back around 12 PM. There was another team that would go up
around 4 PM and come down around 2 AM. But basically 5 nights of 16
hours is what we did this year. Do you want to add anything to that?>>It’s a great question. We’re limited by the time. But what I will say is that
we reserved four nights. And at the end of four nights,
we tallied up some of the lost time. Because it’s inevitable,
some of the recorders don’t work. Sometimes one of the telescopes
has a technical problem. Something can freeze up, and
you lose a little bit of time. And we were really fortunate, because I
was able to call up some of the directors of the telescopes and
explain the situation. And they wound up giving us
some extra time to do that. So to make up for the lost time and
to give us the full exposure, if you will,
that the Event Horizon Telescope wanted. And that was really wonderful to get. And it made all the difference for
the last day. It could have been our best day,
best night.>>Next question in
the front here to your left.>>Why is it so crucial to prove
Einstein’s theory near black holes?>>Can I please take that one?>>Yeah.>>What a wonderful question that is. So Feynman said that the test of
all theory is experiment, right? I mean, you always want to push The
theories that you rely on to the extreme. You always wanna see where they break. So for hundreds of years,
Newton’s theory of gravity was fine. It governed the trajectory of cannonball
shells, or dropping apples and things like that. It’s only when you pushed it and you looked to things like the perihelion
of mercury or other things that were not really understood that you saw
that there was a problem, right? So where are you gonna find a problem? You go to laboratories of extremes that
are given to you by nature, right? You go to neutron stars,
where magnetic fields are insanely more strong than they are here on earth,
right, to test certain things? And where do you go to test
Einstein’s theory of gravity? Where gravity is the dominant
force at the edge of a black hole. That’s where we may seem some
deviation from Einstein’s theory. And it’s important, I wanna say this,
thiis very important to say, for me, if you had asked Einstein
what good his theory was in 1915, what could he have possibly said, right? I mean, there were horse-drawn
carriages in the streets, right? There were no refrigerators,
it was a different time. And here he was thinking about
the geometry of space time. He would’ve said, there’s absolutely
no application I can see. And yet we use it every day today. Without Einstein’s corrections to gravity,
GPS doesn’t work, right? If you didn’t make the Einstein general
relativity corrections then GPS would be off by miles. So sometimes you’ve got to wait 100
years to see if something is useful, but it’s worth it.>>Great, we have the next
question right here in the front.>>Thanks again for
the dual presentations. I would like to ask a question
possible future actions. And this is based on the fact that
according to my understanding, all of your observations
are based on our own galaxy. And there’s one black hole in our galaxy,
maybe. But I would assume that every
galaxy has a black hole. And they’re about ten to the tenth
galaxies, or ten to the eleventh. And I wanted to know, what is the
potential, not in the immediate future, but arranging for some kind of observatory
equipment to be placed in other galaxies? And would that be expected to yield an
advantage over what we can do working on one galaxy?>>You want me to?>>I don’t-
>>Well, I’ll take this, and maybe Katie wants to expand on it. That’s a fantastic point. And what I would say is that the Event
Horizon Telescope is a visionary project. And ten years ago, I would have been
hard-pressed to say that we could actually pull this off, and now I think
we’re quite close to doing just that. And now that we’re close to doing that, we’re thinking exactly along
the lines that you’ve described. What’s next, what’s the next step? Yes, we’re looking at the black
hole at the center of our galaxy. Yes, we’re looking at M87
in the center of Virgo A, which is a thousand times more massive,
right? So it’s a different beast altogether,
right, because it’s in a different galaxy. And the next step is to increase
our magnifying power even more. All right, we’re totally
unsatisfied as scientists, right? So one of the ways we can do that is
by increasing the frequency at which we observe. That’s really hard. You need to develop new electronics,
more stable atomic clocks. So there’s definitely a path forward to
increase your magnifying power just by increasing the frequency. And then we’re excited to think now about
putting a telescope in space, right? How do you leave the surface of the earth,
right? Everything that we’ve both said gives you
a magnifying power if you’re limited to the surface of the earth, but
what if you put something in orbit? Then the baseline becomes from the surface
of the earth to this other spacecraft. Now you’re looking at something that
has a much bigger lever arm, and you can see even finer structures. That will likely bring other
sources into range for doing the kind of work that we described.>>Yeah I’d just like to comment. I think you should be optimistic
because from my limited knowledge of the history of science, major technical
advances are being made regularly and without great time intervals between them.>>I concur.
>>[LAUGH]>>Last question, in the front row to your left.>>Who is funding this project and is there any chance that the current
administration could cut your funds?>>Is this being recorded?>>[LAUGH]
>>Well, so, first of all, it’s a real delight and
a pleasure to report that both Katie and I receive funding from
the National Science Foundation, which has backed this project for
a number of years. We have to submit funding
proposals each year. And if they’re found with merit
then we receive the funding. So we’re very responsible and
excellent stewards of the federal dollar. I really do feel that way. And we don’t go to
the National Science Foundation to ask for the next trials funding until we
have something to show for it. That’s really how we roll in this project. That being said, we’ve also been very, very fortunate to get support from
the John Templeton Foundation and also from the Gordon and
Betty Moore Foundation. And they’ve been instrumental. Because they’ve backed us to
do the kind of visionary, out-of-the-box thinking
that is paving the way for the future of the Event Horizon Telescope,
and to get the job done now. It’s a combination of
this private funding, two very generous foundations and
benefactors, and the National Science Foundation
that’s brought us to where we are now. I really have no comment on the current
administration and whether or not our funding might be cut, other than the fact that I think
we’re just doing really good stuff. And I think that if you’re doing really
good stuff and you have a good team, that shows and people will invest in it.>>[APPLAUSE]
>>Well, that was fascinating and
really wonderful to hear. Intellectual freedom, science crossing
borders and eliminating boundaries. And we only have to wait another year
to find out if there is a snowman->>[LAUGH]>>Or an elephant, or something else and what it looks like. So it’s very exciting, your work. And thank you so much for
coming to share with us where you are. We’ll have to have you back in a year so
we can see the other guests.>>[APPLAUSE]
>>Thank you.>>Thank you, thank you very much.

Leave a Reply

Your email address will not be published. Required fields are marked *