Ralf Vandebergh's tweet shows his own reconstruction of an image of the accretion disk around a black hole, generated presumably from Event Horizon Telescope data.

I'm assuming that the data came from a publicly available source, but I don't know for sure. That would make sense, they used two completely independent groups to reconstruct the images to cross-check each other before they were published, because with so few of the telescopes in the array recording data at any one time, the reconstruction took a bit of artistic license or "filling in".

Since the raw data is huge, boxes and boxes of hard drives with high speed digital recording of GHz rate ADC conversions, I don't think there's an ftp site for the actual raw data. And since calibrations and baselining is pretty complicated, it would be hard to reconstruct.

So I'm wondering what kind of data is available that would allow someone to reconstruct their own EHT image?

FYI Ralf Vandebergh's astrophotography can be found in links within Why exactly do “CMOS astrophotographers” prefer CMOS sensors?

From tweet:

This is one of my processings of the #blackhole #EHTBlackHole first ever image. @RHdijkgraaf @govertschilling @NASAblueshift @ehtelescope @jordydavelaar @UofA @perimeter @smithsonian @TheNRAO @PhysicsWorld

enter image description here click for larger

  • 3
    $\begingroup$ My reading of the tweet is that he did image processing. He didn't do reconstruction. $\endgroup$
    – James K
    Jul 7, 2019 at 9:21
  • $\begingroup$ @JamesK I think it may be something like that. Do you think he just took the image released to the public and processed it further? In that case it's not particularly interesting at all. Everyone's computer has a "sharpen" feature somewhere. $\endgroup$
    – uhoh
    Jul 7, 2019 at 14:09
  • 3
    $\begingroup$ If you look at the (tiny) annotations at the bottom, the left one says, "Original", and the right one says, "Processing Ralf Vandebergh". So I agree with James K -- he just applied some tweaks to the publicly released color image. $\endgroup$ Jul 7, 2019 at 17:51
  • 2
    $\begingroup$ For what it's worth, you can find links to at least some of the pipeline-processed VLBI data here: eventhorizontelescope.org/for-astronomers/data $\endgroup$ Jul 7, 2019 at 17:54
  • 1
    $\begingroup$ @PeterErwin oh that looks interesting! I wonder if I start clicking on those links if I'll be able to get anything else done this week, or if my productivity will fall into an event horizon as well? $\endgroup$
    – uhoh
    Jul 7, 2019 at 21:04

1 Answer 1


From whence is the Event Horizon Telescope black hole data available for amateur reconstruction?

So I'm wondering what kind of data is available that would allow someone to reconstruct their own EHT image?

In the CalTech video Imaging the Unseen: Taking the First Picture of a Black Hole - Katie Bouman - 6/7/2019 which I just found in this question, after about 25:45 Dr. Katie Bouman says (edited from closed captions):

So we collected five petabytes of data during that time... and a little less than a petabyte was for M87... but actually what we use for imaging; we don’t work with the petabytes of data.

This is reduced down through a process of correlation and a calibration stage that initially it’s at petabytes, then it goes to terabytes, then eventually what we use for imaging is in the megabytes range, where you can just email the file. And you can actually download it online now. We have all of the data online that we use for imaging and you can try to make your own image...

It seems that the MB sized raw data is available here, found on the EHT website here per @PeterErwin's comment.

More information is available at https://github.com/eventhorizontelescope/2019-D01-02

One can only speculate on how the tweeted image was generated, but luckily that wasn't the question.

A sample of a random selection of the text files begins with:

# SRC:M87 DATE(MJD):57849 FREQ:227.0707GHz
# time(UTC)      T1     T2   U(lambda)        V(lambda)         Iamp(Jy)    Iphase(d)   Isigma(Jy)
00.86805555      AA     PV  -4396973056.0000  -4510950912.0000  0.13778746    178.1494  0.00382148
00.87083334      AA     PV  -4397497344.0000  -4511639040.0000  0.13610664   -179.6852  0.00413867
00.87361112      AA     PV  -4398009344.0000  -4512314368.0000  0.13498885    179.0204  0.00400464
00.87638891      AA     PV  -4398538752.0000  -4513015296.0000  0.13268795    179.9253  0.00369624

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .