we get an up-close look at a supermassive black hole.

在「星際效應」這部電影中，

Set against a backdrop of bright gas,

我們更近距離地看到了 超質量黑洞。

the black hole's massive gravitational pull

在明亮氣體的背景下，

bends light into a ring.

黑洞的巨大引力

However, this isn't a real photograph,

使光線形成戒指般的環狀。

but a computer graphic rendering --

但是，這不是張真實的照片，

an artistic interpretation of what a black hole might look like.

而是電腦圖像的呈現，

A hundred years ago,

是對黑洞可能的呈像的 藝術化的演繹。

Albert Einstein first published his theory of general relativity.

一百年前，

In the years since then,

愛因斯坦首先發表了他的相對論。

scientists have provided a lot of evidence in support of it.

那之後的幾年，

But one thing predicted from this theory, black holes,

科學家提供了很多證據支持他的理論。

still have not been directly observed.

但是從他的理論中預測到的黑洞

Although we have some idea as to what a black hole might look like,

仍然無法有直接證據證實。

we've never actually taken a picture of one before.

雖然我們對於黑洞的呈像有一些想法，

However, you might be surprised to know that that may soon change.

但是我們從來沒有真正 拍攝過一張黑洞的相片。

We may be seeing our first picture of a black hole in the next couple years.

也許你會驚訝於這種困境即將改變。

Getting this first picture will come down to an international team of scientists,

我們在未來幾年內也許 可以得到第一張黑洞的相片。

an Earth-sized telescope

國際的科學家團隊 將會獲得這第一張圖片，

and an algorithm that puts together the final picture.

透過地球大小般的望遠鏡，

Although I won't be able to show you a real picture of a black hole today,

和一個演算方法，獲得最後這張圖片。

I'd like to give you a brief glimpse into the effort involved

雖然，我今天無法讓大家看到 黑洞真正的照片，

in getting that first picture.

但是，我想要簡單地向各位說明一下

My name is Katie Bouman,

獲得這首張照片所付出的努力。

and I'm a PhD student at MIT.

我是 Katie Bouman ，

I do research in a computer science lab

一名麻省理工學院博士生。

that works on making computers see through images and video.

我在電腦科學實驗室做研究，

But although I'm not an astronomer,

讓電腦透過影像及影片， 能夠「看見」、識別。

today I'd like to show you

雖然我不是天文學家，

how I've been able to contribute to this exciting project.

但是，我現在要讓大家看的是

If you go out past the bright city lights tonight,

我如何投入這令人興奮的專案。

you may just be lucky enough to see a stunning view

如果今晚你們離開了城市明亮的燈光，

of the Milky Way Galaxy.

可能運氣夠好，

And if you could zoom past millions of stars,

可以看到銀河系美麗的影像。

26,000 light-years toward the heart of the spiraling Milky Way,

如果你的視野能夠穿越數百萬顆星星，

we'd eventually reach a cluster of stars right at the center.

向着銀河的螺旋中心前進 26,000 光年，

Peering past all the galactic dust with infrared telescopes,

最後會在中心點遇到一群星星。

astronomers have watched these stars for over 16 years.

天文學家使用紅外線望遠鏡 透過銀河系塵埃

But it's what they don't see that is the most spectacular.

觀察這些星星， 已經超過了 16 年。

These stars seem to orbit an invisible object.

但是，最為壯觀的東西， 卻是他們無法看見的。

By tracking the paths of these stars,

這些星星似乎繞著一個 隱形的物體運轉著。

astronomers have concluded

藉由追蹤這些星星的軌跡，

that the only thing small and heavy enough to cause this motion

天文學家得到一個結論：

is a supermassive black hole --

只有一個又小又重的物體 才能夠造成這樣的運動軌跡，

an object so dense that it sucks up anything that ventures too close --

那就是超質量黑洞，

even light.

它的密度高到能夠吸收 所有敢於近距離靠近它的東西，

But what happens if we were to zoom in even further?

連光線也不例外。

Is it possible to see something that, by definition, is impossible to see?

但是，如果我們將影像放大， 會發生什麼事呢？

Well, it turns out that if we were to zoom in at radio wavelengths,

有沒有可能看到那些 原本被定義為看不見的東西呢？

we'd expect to see a ring of light

事實顯示，如果我們 以無線電波長的尺度放大，

caused by the gravitational lensing of hot plasma

我們預期可以看到一個光環，

zipping around the black hole.

它是由黑洞旁 高速移動的熱離子體的

In other words,

「引力透鏡」效應形成。

the black hole casts a shadow on this backdrop of bright material,

換句話說，

carving out a sphere of darkness.

黑洞在明亮物質的背景下投射出陰影，

This bright ring reveals the black hole's event horizon,

刻畫出黑色的球體。

where the gravitational pull becomes so great

這個光環揭露了黑洞的表面界限，

that not even light can escape.

在那個地方，引力拉扯的力量很大，

Einstein's equations predict the size and shape of this ring,

連光線都無法逃脫。

so taking a picture of it wouldn't only be really cool,

愛因斯坦方程式預測了 這個光環的大小與形狀，

it would also help to verify that these equations hold

所以拍攝黑洞的相片不只是很酷，

in the extreme conditions around the black hole.

它也有助於驗證這些方程式

However, this black hole is so far away from us,

能在黑洞附近這樣的 極端環境下成立。

that from Earth, this ring appears incredibly small --

但是，這個黑洞距離我們非常遙遠，

the same size to us as an orange on the surface of the moon.

從地球看過去， 這個光環是不可思議的小，

That makes taking a picture of it extremely difficult.

就像是月球表面的 一個橘子那樣的小。

Why is that?

所以拍攝黑洞的相片是極其困難的。

Well, it all comes down to a simple equation.

為什麼呢？

Due to a phenomenon called diffraction,

因為，這所有的一切 可以歸結於一個簡單的方程式。

there are fundamental limits

由於「衍射」現象，

to the smallest objects that we can possibly see.

我們所能觀察到的最小物體，

This governing equation says that in order to see smaller and smaller,

是有限小的， 我們無法洞察更小的結構。

we need to make our telescope bigger and bigger.

這個方程式說， 為了要看到越來越小的物體，

But even with the most powerful optical telescopes here on Earth,

我們必須製作越來越大的望遠鏡。

we can't even get close to the resolution necessary

但是，即使透過地球上 最強大的光學望遠鏡，

to image on the surface of the moon.

我們還是遠遠達不到

In fact, here I show one of the highest resolution images ever taken

拍攝月球表面的影像所需要的解析度。

of the moon from Earth.

事實上，請大家看看這張從地球拍攝的

It contains roughly 13,000 pixels,

解析度最高的月球照片之一，

and yet each pixel would contain over 1.5 million oranges.

這張相片大約有一萬三千像素，

So how big of a telescope do we need

而每一個像素可包含 超過 150 萬個橘子。

in order to see an orange on the surface of the moon

那麼，為了要看到月球表面的橘子，

and, by extension, our black hole?

我們需要多大的望遠鏡呢？

Well, it turns out that by crunching the numbers,

再者，為了要看到黑洞？

you can easily calculate that we would need a telescope

事實證明，透過大量運算，

the size of the entire Earth.

我們可以很容易地計算出 我們所需要的望遠鏡

(Laughter)

必須是整個地球那麼大。

If we could build this Earth-sized telescope,

(笑聲)

we could just start to make out that distinctive ring of light

如果我們建造出地球般大小的望遠鏡，

indicative of the black hole's event horizon.

我們馬上就可以探測出一個獨特光環，

Although this picture wouldn't contain all the detail we see

它表明了黑洞的表面界限。

in computer graphic renderings,

雖然這張相片沒有包含所有細節，

it would allow us to safely get our first glimpse

像我們在電腦圖形渲染上看到的那樣，

of the immediate environment around a black hole.

但是，至少我們可以安全地

However, as you can imagine,

對黑洞附近的環境瞥上一眼。

building a single-dish telescope the size of the Earth is impossible.

然而，如同大家想像的，

But in the famous words of Mick Jagger,

建造一個地球大小的 單碟望遠鏡是不可能的。

"You can't always get what you want,

但是在 Mick Jagger 的名言中：

but if you try sometimes, you just might find

「你無法一直得到你所想要的，

you get what you need."

但是如果你去嘗試，你可能會發現

And by connecting telescopes from around the world,

你得到了你所需要的。」

an international collaboration called the Event Horizon Telescope

藉由連結世界各地的望遠鏡，

is creating a computational telescope the size of the Earth,

名為「事件視界望遠鏡」的國際組織

capable of resolving structure

正著手創建一個地球大小的 計算型望遠鏡，

on the scale of a black hole's event horizon.

它能夠解析黑洞的

This network of telescopes is scheduled to take its very first picture

表面界限的結構。

of a black hole next year.

這個望遠鏡網路預計明年

Each telescope in the worldwide network works together.

拍攝黑洞的第一張相片。

Linked through the precise timing of atomic clocks,

世界各地的望遠鏡網路同時運作。

teams of researchers at each of the sites freeze light

透過原子鐘的精準時間鏈結，

by collecting thousands of terabytes of data.

每個地點的研究團隊 藉由蒐集數千兆兆字節的數據

This data is then processed in a lab right here in Massachusetts.

將光線「定格」。

So how does this even work?

麻薩諸塞州這裡的實驗室 接下來處理這些資料。

Remember if we want to see the black hole in the center of our galaxy,

那麼，這些資料是如何運作的呢？

we need to build this impossibly large Earth-sized telescope?

還記得嗎？如果我們想要 看清在銀河中間的黑洞，

For just a second, let's pretend we could build

我們就需要建造地球大小的 望遠鏡，這是不現實的。

a telescope the size of the Earth.

等一下，假設我們能夠建造

This would be a little bit like turning the Earth

地球般大小的望遠鏡。

into a giant spinning disco ball.

就有點像將地球

Each individual mirror would collect light

想像成舞廳裡的迪斯可旋轉球。

that we could then combine together to make a picture.

每一面鏡子會蒐集光線，

However, now let's say we remove most of those mirrors

然後我們能將這些 影像整合成一張圖片。

so only a few remained.

但是，現在讓我們 移除大多數的鏡子，

We could still try to combine this information together,

只剩下少數幾個。

but now there are a lot of holes.

我們仍可試著整合這些資訊，

These remaining mirrors represent the locations where we have telescopes.

但是，現在只能看到很多「孔洞」。

This is an incredibly small number of measurements to make a picture from.

這些剩下的鏡子代表 那些有望遠鏡的地方。

But although we only collect light at a few telescope locations,

測量數據少之又少， 甚至無法形成一張圖片。

as the Earth rotates, we get to see other new measurements.

雖然我們只在少數 有望遠鏡的地方蒐集光線，

In other words, as the disco ball spins, those mirrors change locations

地球旋轉時，我們可以 獲得一些新的測量數據。

and we get to observe different parts of the image.

換句話說，就像迪斯可球旋轉時， 那些鏡子也會改變位置，

The imaging algorithms we develop fill in the missing gaps of the disco ball

我們得以觀察不同面向的影像。

in order to reconstruct the underlying black hole image.

我們所開發的成像算法填補了 「迪斯可球」的不可見縫隙，

If we had telescopes located everywhere on the globe --

目的在重建黑洞的相片。

in other words, the entire disco ball --

如果地球的每個地方都有望遠鏡，

this would be trivial.

也就是整個迪斯可球佈滿了鏡子，

However, we only see a few samples, and for that reason,

這是最簡潔、理想的情況。

there are an infinite number of possible images

但是，我們只看得到 某些局部的成像，因此，

that are perfectly consistent with our telescope measurements.

有無數可能的相片

However, not all images are created equal.

可以與現有望遠鏡的 局部成像相吻合。

Some of those images look more like what we think of as images than others.

當然，並不是每一張「相片」的 優先級別都相同。

And so, my role in helping to take the first image of a black hole

有些相片比別的 更近似我們所想像的。

is to design algorithms that find the most reasonable image

因此，為了協助拍攝黑洞 第一張相片，我的任務就是

that also fits the telescope measurements.

設計發現最合理影像的演算法，

Just as a forensic sketch artist uses limited descriptions

當然也必須符合望遠鏡的量測數據。

to piece together a picture using their knowledge of face structure,

就像法庭的素描家一樣， 利用有限的相貌描述以及

the imaging algorithms I develop use our limited telescope data

他們對於臉部結構的知識， 將表現相貌特點的圖片拼湊出來，

to guide us to a picture that also looks like stuff in our universe.

我開發的影像演算法 使用有限的望遠鏡資料

Using these algorithms, we're able to piece together pictures

為我們生成這種影像： 類似於宇宙中的事物的影像。

from this sparse, noisy data.

利用這些演算法， 讓我們能夠利用零零散散的資料

So here I show a sample reconstruction done using simulated data,

拼湊出黑洞可能的樣子。

when we pretend to point our telescopes

在這裡，讓大家看一個利用模擬資料 重建的影像樣本，

to the black hole in the center of our galaxy.

這是我們假設將望遠鏡指向

Although this is just a simulation, reconstruction such as this give us hope

銀河系中心的黑洞時所得到的。

that we'll soon be able to reliably take the first image of a black hole

雖然這只是一個模擬， 但是這讓我們充滿了希望：

and from it, determine the size of its ring.

我們能夠仰賴這樣的模擬演算法， 很快地得到黑洞的第一張相片，

Although I'd love to go on about all the details of this algorithm,

同時也能計算「光環」的大小。

luckily for you, I don't have the time.

雖然我很樂意繼續說明 這個演算法的所有細節，

But I'd still like to give you a brief idea

但由於時間不夠，所以 你們也不用費腦子聽了。

of how we define what our universe looks like,

但是，我還是很樂意 跟大家做個簡短的說明：

and how we use this to reconstruct and verify our results.

我們如何定義宇宙看起來像什麼？

Since there are an infinite number of possible images

以及我們如何 利用這個演算法重建並驗證結果。

that perfectly explain our telescope measurements,

因為有無數可能的影像

we have to choose between them in some way.

與地球上望遠鏡的量測完全符合，

We do this by ranking the images

我們必須在它們之間 找個方法進行挑選。

based upon how likely they are to be the black hole image,

我們對影像進行打分，

and then choosing the one that's most likely.

打分的根據是：看起來有多像黑洞，

So what do I mean by this exactly?

然後選擇最像的影像。

Let's say we were trying to make a model

那麼，這到底是什麼意思呢？

that told us how likely an image were to appear on Facebook.

假設我們試著建立一個模型，

We'd probably want the model to say

它告訴我們這個影像在 Facebook 上出現的可能性。

it's pretty unlikely that someone would post this noise image on the left,

我們希望這個模型會這樣判斷：

and pretty likely that someone would post a selfie

大家應該不太可能會上傳 像左邊這張亂亂的圖，

like this one on the right.

而比較可能會上傳自拍照，

The image in the middle is blurry,

像右邊這張。

so even though it's more likely we'd see it on Facebook

中間這張圖像片是模糊的，

compared to the noise image,

即使模糊，和亂亂的圖像比較的話， 我們還是很有可能

it's probably less likely we'd see it compared to the selfie.

在 Facebook 上看到，

But when it comes to images from the black hole,

只不過不如自拍照那樣常見。

we're posed with a real conundrum: we've never seen a black hole before.

但是，如果是黑洞的影像，

In that case, what is a likely black hole image,

我們遇到一個真正的難題： 我們從來沒見過黑洞的樣子。

and what should we assume about the structure of black holes?

在這種情況下， 黑洞可能的影像是什麼？

We could try to use images from simulations we've done,

我們應該假設黑洞的結構是什麼？

like the image of the black hole from "Interstellar,"

我們可能會試著使用 之前生成的模擬結果，

but if we did this, it could cause some serious problems.

像「星際效應」裡的黑洞影像，

What would happen if Einstein's theories didn't hold?

但是，如果這樣做的話， 可能會造成一些嚴重的問題。

We'd still want to reconstruct an accurate picture of what was going on.

如果愛因斯坦的理論不適用的話， 會發生什麼事？

If we bake Einstein's equations too much into our algorithms,

我們還是想要重建 一個準確的圖像。

we'll just end up seeing what we expect to see.

如果將太多愛因斯坦的方程式 融入我們的演算法中，

In other words, we want to leave the option open

最後只會得到我們期望的結果， 而不一定是事實。

for there being a giant elephant at the center of our galaxy.

換句話說，我們不能 貿然確定實際情況如何，

(Laughter)

因為銀河系中央有一隻巨象。

Different types of images have very distinct features.

(笑聲)

We can easily tell the difference between black hole simulation images

不同類型的影像有著 各自非常顯著的特徵。

and images we take every day here on Earth.

我們可以很容易地區分 黑洞模擬影像

We need a way to tell our algorithms what images look like

以及我們在地球上日常生活中的照片。

without imposing one type of image's features too much.

我們需要一種方法來告訴演算法 影像看起來像什麼，

One way we can try to get around this

而不是去強加特定一種影像的特徵給它。

is by imposing the features of different kinds of images

我們可以用一個方法 試著解決這個問題：

and seeing how the type of image we assume affects our reconstructions.

通過導入不同類型影像的特徵 讓演算法重建影像，

If all images' types produce a very similar-looking image,

然後觀察預先假設的影像類型 如何影響我們重建的影像。

then we can start to become more confident

如果所有不同類型的影像特徵 產生的結果都很類似，

that the image assumptions we're making are not biasing this picture that much.

那麼我們可以充滿信心地說：

This is a little bit like giving the same description

對於這個影像所做的假設 沒有與事實偏差太多。

to three different sketch artists from all around the world.

這有點像是將相同的相貌描述

If they all produce a very similar-looking face,

提供給三個來自世界各地不同的素描家，

then we can start to become confident

如果他們都畫出很相像的臉，

that they're not imposing their own cultural biases on the drawings.

那麼我們可以充滿信心地說：

One way we can try to impose different image features

他們的作品沒有受到 本人的文化偏見的影響。

is by using pieces of existing images.

我們導入不同類型影像 的特徵的一個方法

So we take a large collection of images,

就是藉由現存的影像去拼湊。

and we break them down into their little image patches.

所以我們要蒐集大量的影像，

We then can treat each image patch a little bit like pieces of a puzzle.

然後將它們分解成許多碎片。

And we use commonly seen puzzle pieces to piece together an image

之後我們可以把這些碎片 當作拼圖的碎片。

that also fits our telescope measurements.

我們使用常見的「碎片」拼湊成圖片，

Different types of images have very distinctive sets of puzzle pieces.

這張圖片當然也要 符合望遠鏡的量測數據。

So what happens when we take the same data

不同類型的影像有其獨特的拼圖碎片。

but we use different sets of puzzle pieces to reconstruct the image?

所以，當我們利用相同的數據資料

Let's first start with black hole image simulation puzzle pieces.

卻使用不同類型的拼圖碎片 來重建這個影像，會發生什麼事？