Hacker News Reader

Hacker News

Comment by piker | Hacker News Reader

Full thread

piker

Higher resolution official release seems to be: https://www.nsf.gov/news/special_reports/blackholes/download...

View on HN

Does that image have a weird pulsing optical illusion for anyone else?

Relieved it isn't just me

I get it if I move a bit when on an angle, I assume it's just my crappy monitor's viewing angles causing it.

Yes, fascinating! If you shift the focus of your eyes from the outside to the inside of the hole, it seems to "pulsate"

Yes, reminds me of the Eye of Sauron.

Funnily, it reminds me of Soundgarden's album, Superunknown, which had the famous song 'Black Hole Sun.'

https://www.amazon.com/Superunknown-Soundgarden/dp/B00IXLQJ8...

The supreme astronomical example of the Eye of Sauron is this Hubble shot of the star Fomalhaut and its debris disk: https://en.wikipedia.org/wiki/Fomalhaut#/media/File:Fomalhau...

It's the color scheme. I think in matplotlib it's called "hot". I do a lot of 2D kernel density plots that by having a black zero level and dozens of contour levels produce a smooth look with aesthetics similarities to this. I used to use that scheme because "heatmaps" but stopped because of the pulsing illusion.

I'm not sure if this image is real color or just lightness value and they used a color scheme for drama.

It's the latter. This is an image in radio-frequency brightness of the object, so not visible frequencies of light. But yeah, it also looks like the 'hot' colormap to me.

Grrr, why aren't scientists all using perceptually uniforms color schemes! This would still look awesome in `plasma`. Is there any where I can get the data and do it myself?

One of my favorite talks ever is on this subject: https://www.youtube.com/watch?v=xAoljeRJ3lU

Can you run it through the colormap in reverse to (approximately) recover the data?

Technically yes -- colormaps are 1-dimensional non-self intersecting curves in RGBA space. You just grab them by the... I mean just straighten them.

I don't know if an explicit formula for the colormap is given, but you can always do "xs = np.linspace(0,1,100); ys = { cm.hot(x):x for x in xs}" and recover an approximate inverse. Then apply this function to each pixel of the image.

Based on my analysis the published image is too processed to be able to invert the colormap to get to the brightness temperature data. I had better luck with Paper IV Fig. 15. See: https://physics.stackexchange.com/questions/472641/what-woul...

'Inferno' would be nice for this, also perceptually uniform, and has the benefit that space is still black.

You can easily remap it between different colour maps if you assume that is is a standard hot map.

Some time ago I've seen a picture over the Internet where the description said the more you see it "moving" the more likely you're stressed.

As for the black hole, looking it at full screen I see it pulsating a little.

I don't even have to move my eyes, I stare at it and it's constantly moving away from me. I had to check that it's not actually a movie. Strangest thing.

It does look like it pulsates to me! That's an interesting illusion :)

You mean to tell me this is not a gif?!

I literally did double-check the file extension

Interesting that so many people see this but for me it's absolutely static.

If you are focused and calm, or have bad sight, you won't see it naturally. To force it, move your focus from 1cm to the left, then right, again and again, quickly.

I suspect this is because it's extremely blurry, if you watch it on a large screen it might trick your eye into trying to refocus, obviously without success.

I think my eye is trying to bring it into focus. It is like when the camera on your phone can't quite figure out what it is looking it and keeps adjusting.

looks like the image data was cleaned up with bicubic interpolation, or something of the like.

https://en.wikipedia.org/wiki/Bicubic_interpolation

Stare at it till there are two, then have fun snapping them in and out.

moving the picture on my screen caused some eye bug

Yes, and I also see a spot in the shape of it after looking away, though that may just mean my monitor is too bright

Yes if you dance your focus around the image, it appears to pulsate. Probably your pupils constricting and dilating due to the high contrast in the image.

And a really great explanation of why it looks the way it does:

https://www.youtube.com/watch?v=zUyH3XhpLTo

editing that image reveals all kinds of details: https://imgur.com/a/TvQsgbA (and I don't even know what I'm doing)

Mind you, that's a JPG you are editing, so in the 8 bit sRGB color space. That means manipulations can lead to errors and illusions due to the color space being non-linear and clipped[0].

The original data was likely linear and at a much higher precision. If the source was a 16 bit linear grayscale PNG for example you could be much more assured you're not seeing the effects of JPG compression and things that were actually measured.

EDIT: Found better sources:

16-bit sRGB PNG: https://eventhorizontelescope.org/files/eht/files/20190410-7...

180 MiB original TIFF: https://www.eso.org/public/images/eso1907a/

[0] https://www.youtube.com/watch?v=LKnqECcg6Gw

Thanks a lot for the explanation and the better sources, the TIFF image should probably be a torrent because it's not downloading very fast, the PNG image does give much nicer renders: https://i.imgur.com/zZcD5Na.jpg

Those 'details' aren't real. The image is the result of a complex interpolation algorithm that takes very noisy and incomplete data as its input. The resolution is quite limited.

I do (I think...) the bands are atifacts:

https://imgur.com/a/dhN9Pf9

The article linking to that image: https://nsf.gov/news/news_summ.jsp?cntn_id=298276

Thanks for the source article! I had heard about this but not seen any material on it until now.

The additional pixel density probably doesn't add any details though. The original image looks like a smoothed-out version of a low-res image anyway...

Is this link down for anyone else?

Probably because of HN effect again. Yes it is down.

Can someone stick it on IFPS?

Slashdot effect

So how much this picture cost ?

$40m

So basically nothing in the grand scheme of anything.

I've taken to measuring moderately large amounts of money in "juiceros". Forty million dollars is one third of a juicero.

That's half the cost of a single F-35 fighter plane. Just US alone will pay for around 2500 of them (2010 estimate), and plans to pay around 4 times more than that to maintain them. So just F-35 total costs for US are equivalent to 25000 scientific projects like this one. To compare, it would be one new project like this every day for 70 years!

This specific project is more European or even a world project than just US, if I understood correctly.

It makes sense to compare the costs to a single plane but to the whole fleet is just stupid. You can't defend a country with scientific projects.

The thing is every time someone proposes the idea to slash the military budget to fund something else there are at least a hundred other people with a different idea on what to use the funds on. If the funds were spread over so many different projects you end up with an insignificant sum in each of them. Spending the money on military might be actually be advantageous because of large investments in new military technology end up benefiting the civilian sector. (Isn't that the point of the F-35?)

It’s the opposite: most of the military money goes to the stuff which, if actually used, ends the human civilisation as we know it.

The same goals that US has could be achieved with orders of magnitude less military spending, while also reducing the risks for the whole humanity.

So every alternative to the current practices is infinitely better.

Whenever this type of question comes up, how much science cost, think about the following:

It costs more to do a sciency Hollywood movie about than it costs to actually do the science. Sending an actual probe to the orbit of Mars is in general cheaper than making a sci-fi movie.

So...

the original tiff is 7416x4320

it's still too many pixel, it can be a fifth of that while conveying the same amount of information.

I'm puzzled by this, if each of those pixel is actually captured by the lenses, why is it all this much uniform? was it smoothed or does this suggest that it's actually a gigantic uniform cloud of gas?

haven't seen the whole video of the release, I'll have to catch up later in the evening, but this really seems to have captured way too much compared to the actual lens resolution and I wonder what would be the "confidence interval" or astrophysical equivalent on each of those pixels.

The picture is not captured by lenses, or an optical telescope at all. It is created by inverting the data received at eight radio telescopes (or eighty individual dishes) around the world. And the smoothing is just due to the inherent limit in the resolving power of the telescope array. The impressive bit is that we see more than a single bright dot.

There is no lens. This was done using eight radio telescopes (or arrays) around the globe to create a "virtual" radio telescope which is effectively the size of the earth. After that the data of the individual telescopes was processed to produce an image.

There is some more info in the wikipedia article for the Event Horizon Telescope (EHT): https://en.wikipedia.org/wiki/Event_Horizon_Telescope

Thanks for the link. More reading reveals the amount of data produced goes into Petabytes. You can't just upload it to drop box or push it via FTP; hence correct icon of an airplane because those drives with data are indeed delivered by regular means of transportation.

https://en.wikipedia.org/wiki/Event_Horizon_Telescope#/media...

From the press conference, they said that the raw captured data took up 6 cubic meters of hard drives

> cubic meters of hard drives

Astronomy and high energy physics are pretty much the only science I know where this is an applicable unit of measurement

Reply on news.ycombinator.com