1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

First pictures of a black hole in space

Discussion in 'off topic' started by gavreid, Apr 10, 2019.

  1. Mullardman

    Mullardman Moderately extreme...

    I bet you didn't..;)
  2. Ragaman

    Ragaman Mentalist

    I'll stick with my theory.
  3. Mr Cat

    Mr Cat Owner

  4. Bob McC

    Bob McC Living the life of Riley

    Moon landing?

    Not quite.
  5. Cav

    Cav pfm Member

    No, I am sure the lander/Moon interface happened....
    Mullardman likes this.
  6. cooky1257

    cooky1257 pfm Member

    Facing the wrong way...
    darrenyeats likes this.
  7. Joe P

    Joe P certified Buffologist / mod



  8. Joe P

    Joe P certified Buffologist / mod

    Sorry, that's a cat's hole in my face, not a black hole in space.

  9. martin clark

    martin clark pinko bodger

    Black Holes can only emit Hawking radiation; cats IME, are capable of far more toxic emissions.
    Joe P likes this.
  10. Jim Audiomisc

    Jim Audiomisc pfm Member

    Having - looong ago - worked on 'mm-wave' kit I had a look at what details I could of the systems they used. Interested to see that the signals they received were at 230GHz and 345GHz. These used to be the 'popular choice' for mm-wave astronomers when we developed the first receivers for them in the 1980s because they cover the main emission line of Carbon Monoxide which is fairly common as a constituent of interstellar clouds, etc. So they seem to have remained the common choice for this range.

    The systems used here, though had a 4GHz bandwidth and each RX's output was recorded at 16Gigasamples/sec. :) Run over 5 nights on all the telescopes used, that tots up to a seriously large bucketload. Nice to see the JCMT was a part of this as I did some of the work on its optics. Just a shame that the UK gave the telescope away because our goverment decided we didn't need it. Idiots.
    gavreid likes this.
  11. droodzilla

    droodzilla pfm Member

    Here's an interesting Facebook post from someone called Misty S. Boyer. Sadly, right-wing trolls are attacking Katie Bouman, and seeking to minimise her contribution to the historic image (of course, most major results like this are a huge collaborative effort, but it's clear that Bouman made a highly significant contribution. Anyway, here's the FB post, which also addresses some concerns, expressed on this forum, that the image is "made up" in some way...

    There's a lot of news going on about the "black hole girl" right now, and how she's being given too much credit for her role in the historic first image of a black hole. Because this is too important, I want to set the record straight.

    Once Katie Bouman became the "face" of the black hole photo, and articles began to call her "the woman behind the black hole photo", an assortment of people that I'm strongly inclined to call incels but won't decided to figure out just how much of a role she had in it. Why? You'd have to ask them. Something about her attractiveness, youthfulness, and femaleness disturbed them to the point where they had to go digging.

    And after digging, they found Andrew Chael, who wrote an algorithm, and put his algorithm online. Andrew Chael worked on the black hole photo as well. And because people kept saying that Katie Bouman wrote "the algorithm", these people decided that "the algorithm" in question must be Chael's.

    So they looked at Chael's GitHub repository and checked the history. The history showed that Andrew Chael's commits totaled more than 850,000 lines, while Katie Bouman contributed only 2,400.

    "Oh my god!" they all said. "He did almost all of the work on the algorithm and yet she's the one getting all of the credit!"

    They dug a little deeper - but not much - and discovered that the algorithm that "ultimately" generated the world-famous photo was created a different man, named Mareki Honma.

    "She's taken the credit from two men!" they gasped. "Feminism and the PC media is destroying everything!"

    There were, of course, those who tried to be kind. "She's always said that this was a team effort," they said. "We don't blame her, we blame the media. She didn't ask to become the poster girl of a team project she barely contributed to."

    Meanwhile, Andrew Chael - a gay man - tweeted in defense of her. He thanked people for congratulating him on the work he'd spent years on but clarified that if they were doing so as a part of a sexist attack on Katie Bouman, they should go away and reconsider their lives. He said that his work couldn't have happened without Katie.

    And it turns out that he was the one who took the viral photo of Bouman, specifically because he didn't want her contributions to be lost to history

    So I decided to find out for myself what Katie Bouman's actual contributions were. As a programmer, I'm well aware that the number of GitHub commits means nothing without context. And Chael himself clarified that the lines being counted in the commits were from automatic commits of large data files. The actual software was made up of 68,000 lines, and though he didn't count how many he did personally (having said he doesn't actually care how much of it he personally authored), someone else assessed that he wrote about 24,000 of those.

    Whether 68,000 or 24,000-- it's more than 2,400 right? Why call it "her" algorithm, then?

    Because there's more than one algorithm being referenced here. These people just don't realize it.

    I'll work my way backward because it's easier to explain that way.

    The photo that everyone is looking at, the world famous black hole photo? It's actually a composite photo. It was generated by an algorithm credited to Mareki Honma. Honma's algorithm, based on MRI technology, is used to "stitch together" photos and fill in the missing pixels by analyzing the surrounding pixels.

    But where did the photos come from that are composited into this photo?

    The photos making up the composite were generated by 4 separate teams, led by Katie Bouman and Andrew Chael, Kazu Akiyama and Sara Issaoun, Shoko Koyama, Jose L. Gomez, and Michael Johnson. Each team was given a copy of the black hole data and isolated from each other. Between the four of them, they used two techniques - an older, traditional one called CLEAN, and a newer one called RML - to generate an image.

    The purpose of this division and isolation of teams was deliberately done to test the accuracy of the black hole data they were all using. If four isolated teams using different algorithms all got similar results, that would indicate that the data itself was accurate.

    And lo, that's exactly what happened. The data wasn't just good, it's the most accurate of its kind. 5 petabytes (millions of billions of bytes) worth of accurate black hole data.

    But where did the data come from?

    Eight radio telescopes around the world trained their attention on the night sky in the direction of this black hole. The black hole is some ungodly distance away, a relative speck amidst billions of celestial bodies. And what the telescopes caught was not only the data of the black hole but the data of everything else as well.

    Data that would need to be sorted.

    Clearly, it's not the sort of thing you can sort by hand. To separate the wheat (one specific black hole's data) from the chaff (literally everything else around and between here and there) required an algorithm that could identify and single it out, calculations that were crunched across 800 CPUs on a 40Gbit/s network. And given that the resulting black hole-specific data was 5 petabytes (hundreds of pounds worth of hard drives!) you can imagine that the original data set was many times larger.

    The algorithm that accomplished this feat was called CHIRP, short for "Continuous High-resolution Image Reconstruction using Patch priors".

    CHIRP was created by Katie Bouman.

    At the age of 23, she knew nothing about black holes. Her field is computer science and artificial intelligence, topics she'd been involved in since high school. She had a theory about the shadows of black holes, and her algorithm was designed to find those shadows. Katie Bouman used a variety of what MIT called "clever algebraic solutions" to overcome the obstacles involved in creating the CHIRP algorithm. And though she had a team working to help her, her name comes first on the peer-reviewed documentation.

    It's called the CHIRP algorithm because that's what she named it. It's the only reason these images could be created, and it's responsible for creating some of the images that were incorporated into the final image. It's the algorithm that made the effort of collecting all that data worth it. Any data analyst can tell you that you can't analyze or visualize data until it's been prepared first. Cleaned up. Narrowed down to the important information.

    That's what Katie Bouman did, and after working as a data analyst for two years with a focus on this exact thing - data transformation - I can tell you it's not easy. It's not easy on the small data sets I worked with, where I could wind up spending a week looking for the patterns in a 68K Excel spreadsheet containing only one month's worth of programming for a single TV station!

    Katie Bouman's 2,400 line contribution to Andrew Chael's work is on top of all of her other work. She spent five years developing and refining the CHIRP algorithm before leading four teams in testing the data created. The data collection phase of this took 10 days in April 2017, when the eight telescopes simultaneously trained their gazes towards the black hole.

    This photo was ultimately created as a way to test Katie Bouman's algorithm for accuracy. MIT says that it's frequently more accurate than similar predecessors. And it is the algorithm that gave us our first direct image of a black hole.

    Around the internet, there are people who have the misperception that Katie Bouman is just the pretty face, a minor contributor to a project where men like Andrew Chael and Mareki Honma deserve the credit. There are people pushing memes and narratives that she's only being given such acclaim because of feminism. And because Katie Bouman refuses to say that this was anything other than a team effort, even the most flattering comments about her still place her contributions to the photo at less-than-equal contribution to others.

    But I'm writing to set the story straight:

    When it is written that Katie Bouman is the woman "behind the black hole photo", it is objectively true. She wasn't the only woman, but her work was crucial to making all of this happen.

    When Andrew Chael says that his software could not have worked without her, he isn't just being a stand-up guy, he's being literal. And there are those who could just as easily say the same about his contribution, or the contributions of many others.

    And while it's true that every one of the 200+ people involved played an important role, Katie Bouman deserves every ounce of superstardom she receives.

    If there must be a face to this project - and there usually is - then why shouldn't it be her, her fingers twined across her lips, her gleeful eyes luminous and wide with awe and joy?


    Thinking on it a little further, I felt I should clarify that I'm not actually trying to downplay Andrew Chael. His imaging algorithm is actually the result of years of effort, a labor of love. Each image that could be composited into the final photo brought with it a unique take on the data, without which the final photo wouldn't have been complete.

    So let's take a moment to celebrate the fact that two of the most integral contributors to the first direct photo of a black hole

    were a woman

    and a gay man.

    2nd Update (LONG!)

    I went to bed at 19 shares on a post I wrote to vent to my FB friends, and now it's over 2K. I guess it's gone viral. That means I have some work to do.

    I'm going to provide a list of the various articles I read to piece this together. When I wrote this, I wasn't trying to write an essay so I didn't put sources in and I didn't ensure that every detail is 100% accurate. So I'm doing that now.

    Any edits I make are mentioned below (apart from spelling/grammar fixes). The resources that led me to write this are listed below. And because I value accuracy, I welcome people to point out mistakes of any kind. I'll make corrections and credit them here.

    Edit: I incorrectly wrote that Bouman worked on the algorithm for 6 years and spent 2 years refining it. This was an accidental mush of facts: She's been working on this project for a total of 6 years (ages 23 to 29). She spent 3 years building CHIRP and 2 years refining it. I've corrected that and included that she led the four teams, as two separate articles mention it.

    Edit: One of the leads for the 4 team project was a man named Jose L Gomez. I added that to the above, after being sent a twitter thread from Xu S. Han. Thank you! Twitter thread here:

    Edit: Thanks to Zoë Barraclough and someone who would prefer not to be named, for messaging me with another couple of edits. As confirmed on Kazu Akiyama's twitter, there were more than four leaders for the four imaging teams. As I find out the names of these co-leaders, I'll incorporate them into the post.

    This is a 2016 MIT article announcing CHIRP. It gives a pretty excellent idea about the magnitude of Bouman's contribution.

    This goes into detail about Katie Bouman's algorithm. It describes how her algorithm differs from normal/traditional interferometric algorithms. This article explains the difficulty she faced in how trying to capture a black hole is like trying to photograph "a grapefruit on the moon." This also explains how Bouman's algorithm made all of this work-- it combines all of the data from the participating telescopes into, in essence, one massive telescope.

    This is a 2016 TEDx talk from Bouman where she describes her work. Note: though I am intentionally focusing on her contributions specifically to defend the attention she's getting, she makes it clear that this was a team effort. She always gives credit to her teammates who work with her. She is full of humility and wonder.

    This is the paper based on Bouman's work, where she's listed as first author. The position of her name is important. While the meaning of being first author can differ in certain fields, I'm basing the 'primary contributor' interpretation on the fact that multiple other articles say she was lead, MIT refers to the algorithm as hers, as well as the fact that she named CHIRP.

    This is Andrew Chael's imaging library available on GitHub. It's where our original "sleuths" discovered that Bouman had contributed very little and assumed that she was stealing the glory from others. NOTE: Andrew Chael didn't make these claims or ask for this sort of attention!

    This is a paper describing Chael's work, which is impressive. Bouman is in the position of last author. Again, the relevance of the author order can differ, but the common significance of 'last author' is either the supervisor or the relative least contribution. In Bouman's paper, the position of last author seemed to indicate supervisor(s) based on the organization hierarchy on the EHT website. In this instance, I interpret Bouman's name being last as her being a minor contributor to Chael's specific work.

    This is the official EHT telescope website. I can't remember what I looked at here, it's in my history. I think I was trying to find out who Bouman's project lead was.

    This is the twitter thread where Chael defends Katie. He explains that he didn't write 850K lines, defends Katie and says that his algorithm couldn't have worked without her, mentions his LGBTQ status, and more. He seems like a great guy.

    This article speaks to some of the other people involved, including the project leader Sheperd Doeleman. This describes the process they went through in creating the black hole image and is where I got the information about how they split the teams into 4, and how the final image is a composite.

    This is the article that talks about CHIRP sorting through a "true mountain" of data, and how that data was passed out to four teams to check for accuracy.

    This article talks about Bouman coming up with a new algorithm to "stitch data across the EHT network" of telescopes, and how she led an elaborate series of tests (splitting the data up across four teams, etc) to verify that the output wasn't the result of a glitch or fluke.

    This article explains Honma's significant role. It describes what Honma's algorithm does and how it was used in this project.

    Here is another article that goes into more detail about Honma and team. He does a great job of explaining how all of the algorithms in question were, in fact, capable of producing accurate images of the black hole, and a part of the task of his algorithm was to verify the accuracy of those generated photos.

    The final link is the document by all 200+ participants. This document is important because it gives such a clear idea of the work that went into this, the fabric of which Bouman is an integral part. While I'm intentionally highlighting her contributions in defense of her, it should be understood that, like with most scientific breakthroughs, there were many unsung heroes:
    Rob998, PsB, Nick_G and 7 others like this.
  12. Ragaman

    Ragaman Mentalist

    Any scientist in this field will not be making square pegs to fit in round holes, the opposite occurs if anything. Science at this level is all about disproving until the only thing left is the truth, as we have here.

    I would trust the true scientific community over any forum dweller here or anywhere else for that matter, where this level of scientific knowledge is needed.
  13. Jim Audiomisc

    Jim Audiomisc pfm Member

    Just want to say "thanks" to Droodzilla for the excellent posting above. Good research, good journalism.
    PsB and Joe P like this.
  14. Joe P

    Joe P certified Buffologist / mod

    Katie Bouman landed a job as an assistant professor at Caltech. Well done!

    darrenyeats and ian123running like this.
  15. Martyn Miles

    Martyn Miles pfm Member

    Yes, I noticed that.
    I wondered if it was a new kind of hole, one with an apostrophe.
  16. darrenyeats

    darrenyeats pfm Member

    Hole punctuation. See what I tried (and failed) to do there?
  17. stephen bennett

    stephen bennett Mr Enigma

    I haven't read any of the actual publications yet (and may have missed this on here) but how come we see an event horizon around a disc when a black hole is spherical?

    I know the image is a construct, so is it just presented this way?

  18. Cav

    Cav pfm Member

    Look at a solar eclipse and you see the same thing; look at the moon and you see a disc not a sphere.
  19. Jim Audiomisc

    Jim Audiomisc pfm Member

    Two factors occurred to me:

    1) I've assumed this is partly an example of what Astronomers call "Limb Brightening". The gas is a hot but translucent surround, so you get a longer line-of-sight path at the sides than when looked directly into nearer the hole.

    2) (possibly) because we are seeing the hole from above the plane of its galaxy. The hole will presumably be rotating about an axis similar to that of its galaxy given that it is taking up the rotational momentum of the starts, etc, it eats. Near to the hole the infalling matter may tend to be drawing into the plane of rotation in a way similar to how the rings of saturn tend to arrange the fragments into a plane.

    However I'm not an astronomer.
  20. Jim Audiomisc

    Jim Audiomisc pfm Member

    That's a bit different because the moon is between you and the central part. However if you observe at wavelengths where the plasma near the Sun is brighter than the Sun's surface (sic) you can get the limb brightening effect and it would look like a ring despite the plasma surrounding the Sun, and without the Moon in the way.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice