Vera Rubin's primary mirror gets its first reflective coating | Hacker News Reader

Vera Rubin's primary mirror gets its first reflective coating

universetoday.com

157 pointsbikenaga55 comments

From this page[1] linked from the article: "The Rubin Observatory will produce, on average, 15 terabytes of LSST data per night, yielding an uncompressed data set of 200 petabytes. Dedicated facilities will process the image data in near real time."

And "The camera’s 3.2-gigapixel focal plane array comprises 189 4Kx4K CCD sensors with 10 µm pixels. The sensors are deep depletion, back-illuminated devices with a highly segmented architecture that enables the entire array to be read out in two seconds."

Wow that is a huge amount of data from some amazing hardware. This is going to be amazing!

[1] https://kipac.stanford.edu/research/projects/vera-rubin-obse...

Vera Rubin was quite an astronomer [1] both in her discoveries like the rotation of galaxies (confirming dark matter), and for being a trailblazer for women in astronomy. I love reading bios of people like this. Glad to see her get recognized with such a great observatory [2] continuing her legacy of studying dark matter.

[1] https://en.wikipedia.org/wiki/Vera_Rubin

[2] https://en.wikipedia.org/wiki/Vera_C._Rubin_Observatory

Fritz Zwicky discovered Dark Matter. Rubin thought she discovered Dark Matter, because she wasn't aware of Zwicky's paper written in German many years before Rubin's paper.

edit: I have a PhD in Astronomy, with a focus on Dark Matter, and I measured rotation curves of galaxies myself.

Wikipedia link for those who, like me, where unfamiliar: https://en.wikipedia.org/wiki/Fritz_Zwicky

I have not read anywhere that she discovered Dark Matter but rather that her observations confirmed its presence as “first proposed by Fritz Zwicky in 1933 but rejected by most scientists.”

[1] https://www.nsf.gov/news/special_reports/medalofscience50/ru...

"The VRO’s design allows the camera to capture a large area of sky the size of 7 full moons across in a single image."

Wow, that's a lot of sky for a telescope. Around 3.5° at once. I purchased a telescope specifically to view DSO, and it is ~3 full moons. When attaching my older DSLR, a full moon isn't even 50% of the captured image. Of course, mine's only ~150mm instead of 8.4m. Just putting size in perspective of my own use instead of using school buses or football pitches or basketball type nonsense.

It's about the same angle of view that a 400mm lens on a full frame camera has vertically. (wide field astrophotography can be fun: https://nt4tn.net/astro/horse4.jpg )

VRO has a 10313mm focal length (vs 8.4m diameter, so f/1.234-- though with a huge central obstruction) illuminating a 630mm focal plane.

That angle of view corresponds to a lens with 588mm focal length on a full frame camera (horizontally), and 392mm (vertically). Funny that astronomers and photographers mean very different things by saying wide angle. ;-)

People get stuck into trying to apply A to B when B != A. Just because the same words are used does not mean they have the same use and meaning as applied.

Also, "space is big. You just won't believe how vastly, hugely, mind-bogglingly big it is."

It's a full frame equivalent. I don't think anyone is trying to use this on a full frame camera.

Well, if you never compare A to B, you won't believe it for sure. :-/

VRO is a survey telescope. It will survey the all sky visible in the location in just few nights.

It's so, so cool that humanity builds these kinds of things. I'm so grateful that there are people who dedicate their lives to this.

I wish that I had.

you can always switch careers. you'll just have to switch the lifestyle to which you have become accustomed as well. I'm guessing the average bit banger around these parts earns a much higher salary than the average person in astronomy. You do however have the potential of actually helping humanity understand things rather than droning away on ad tech and social platforms which is definitely not pushing humanity in any thing resembling a positive direction.

When I [briefly] looked into this, it seemed that positions in astronomy were extremely competitive, extremely limited, and generally had a requirement of PhD and the usual toppings alongside. Love to know if I was wrong somehow? Astronomy has always been a passion, but I never pursued it because of certain irritating things in cosmology that left me somewhat jaded.

Same. I'm 44. No chance I'm going to be able to have a career in astronomy now.

Not an astronomer but have an advanced degree in physical science, very experienced with optics and engineering, and incredibly interested in the science. I looked into technical positions at a couple of giant telescope projects (not VRO) and got nowhere. I did speak with someone informally about it and was told that the science community was pretty decoupled from the instrument and facility builders.

Really depends what you mean by “a career in astronomy”.

Being a research astrophysicist at a university or major lab likely requires a PhD and (to achieve tenure) many years of postdoctoral experience.

However, systems like Rubin (which I have worked on) require complex data processing and management systems to make them effective. Building those doesn't require expertise a PhD astronomy — although some level of interest and enthusiasm certainly helps — but rather the sorts of engineering skills that the typical Hacker News reader might possess. Skillsets like that are increasingly vital as astronomy moves towards large-scale, data-intensive infrastructures like Rubin, SKA, etc.

That said, it's certainly true that taking your career down this path isn't likely to be as well remunerated as a career in commercial software development (although that varies a bit with geography).

Should you be interested, check us out at https://www.werkenbijastron.nl/ (probably mainly of interest to folks in Europe) or try the American Astronomical Society Job Register at https://aas.org/jobregister?f%5B0%5D=category%3A514&f%5B1%5D....

Very cool! HD video: https://rubin.canto.com/v/gallery/album/HDSNU?display=curate...

The AAT has a pretty good video on aluminising with a lot of exposition, https://www.youtube.com/watch?v=fztRqJerfOk

VRO's process is considerably more sophisticated, as they do the silicon nitride protective coating. I believe the specific approach on VRO is similar to the one pioneered for gemini: https://www.gemini.edu/files/docman/websplash/websplash2004-...

Fancier protected metal (and dialectic enhanced) coatings have been common in smaller reflectors... but it's quite a big difference doing a fancy sputter coating on a 150mm object vs a 8m one! :P

That video was pretty wild. I didn’t understand at first what had happened when the silver layer got laid down and it was suddenly reflective

The video was very cool. The arm swings around and hello! Just like that, it's a shiny mirror. Very impressive effect to be sure

That was breathtaking.

God the work that went into that one shot. Not just camera work, but all the engineering for the mirror. It's astounding.

The mirror is the easy part. In fact, it was finished years ago - just not coated.

lulz. Easy. Okay. It might be the easiest part of this build (I have my doubts to the accuracy of that), but let's not get carried away with slinging words like "easy" around things that are measured to microns accuracy and took years to build.

Maybe not “easy” but it was straightforward in comparison to much of the other things on the project. It was done in the U of A mirror lab and it’s a pretty well-defined process to build these mirrors, with a small exception for this mirror being that it’s a combination M1/M3 with two different grinds, but it’s about as close to a factory process as you can get for this project. Before the mirror lab, Corning would often build the big mirrors, both for science and defense.

The 300T telescope movement assembly which can move the 300+ Ton telescope 3.5 degrees in 5 seconds, and the 3.2 GP camera especially are quite a bit harder to build, engineer, and integrate.

Here's how they cast an 8m class mirror at the University of Arizona. This is for the Giant Magellan Telescope, not the VRO, but it also did that one. \

https://www.youtube.com/watch?v=c-lBKuHqHk0

Grinding and polishing is its own art.

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

None of this looks easy to me.

This telescope and its mission to catch changes in the sky over time is going to help discover so many things.

The reaction will be interesting.

I love that we're also remembering to take less scoped in to see a wider view. I wonder how many times they will complete the full sky survey. Will there be enough to provide a timelapse from multiple surveys, or will it take the entire scope to complete the full sky?

I find some of the timelapse of various objects quite fascinating now that we've been recording observations long enough. Seeing the movement around SagA* is amazing. Just the other day, we saw Cassiopeia A over a couple of decades. Some part of my brain knows things are constantly changing in the universe, but at the time scale it seems strange things are noticeably different within our own lifetimes.

They will survey the entire sky every week or so. At the end, you can create a nice 60 fps timelapse of many objects in the southern sky.

This telescope has a lot of interesting details. The CCD is massive! The largest ever made, and it's going to be downloading so much data it's actually insane. Way more than can be processed by a human.

Wait. I thought we were doing large diameter telescopes as a bunch of individual, tractably-sized mirrors pointing at a common collector. When the idea was presented to me they made it sound like it was impossible to make larger monolithic mirrors because we’d hit a wall.

How did monolithic mirrors get out ahead again?

Answering my own question:

Rubin will not be the largest aperture telescope in the world. It’s 8.4m, while Keck is 10m and no longer the largest segmented mirror. Rubin will be the largest single mirror main reflector, beating 8.2m for the previous largest telescopes.

So it’s incremental improvement not retaking a crown.

8.x m is about the largest single mirror that can be cast with current technology. Above that, segmented wins. Giant Magellan[0] will use multiple 8m mirrors to get to ~25m total aperture. The Thirty Meter Telescope[0] and the European Extremely Large Telescope (39m) will be segmented.

The VRO picked the single mirror approach for lower complexity and the ability to lock the primary and tertiary mirrors into one monolithic structure. For this, they accepted some higher fabrication risk.

Source: A friend on the VRO project.

[0] GMT and TMT are at budget risk and TMT faces stiff opposition from Hawaiians who say Mauna Kea is sacred and shouldn't have anything at the summit.

manufacturing challenges aside, isn't there a practical limit to what terrestrial optical telescopes can achieve, given the interference from our atmosphere? I thought things like JWST was the necessary next step in improvement.

I’ve been binging Goldratt lately, so my brain is awash in Theory of Constraints.

It’s okay to use “old” equipment to offload from a bottleneck. Space based telescopes are going to be a bottleneck for the foreseeable future. They have severe maintenance restrictions and the more you task swap them the more frequent that maintenance.

So from a constraint perspective, you use ground based observation to do anything that can be done without a space telescope. Including trying to determine which patches of sky justify time on JWT.

JWST was conceived ~30 years ago and there's been a lot of advancement since then. I agree that the atmosphere and the background light from the earth and sky can get in the way of terrestrial telescopes. On the other hand, the ability to build giant apertures AND equally enormous instruments have benefits of thri own. Check out the proposed instruments for the 39m E-ELT at

https://elt.eso.org/instrument/

These things are immense. The METIS is supposed to cover 3-13 microns, an impressive accomplishment through the atmosphere. On top of that they can be upgraded and the data can be offloaded with current technology vs. waiting for lasercomm to eventually bring the bandwidth. Even the Hubble is ultimately limited by the chassis. E-ELT expects to do 15x Hubble's resolution and directly characterize exoplanet atmospheres.

Of course if the sky gets gunked up by infinite comsats, all bets could be off.

Dumb question, but why don’t they just use a single more highly-curved mirror rather than a series of 3 bounces?

More curved means more optical aberrations. The Rubin is a so-called three-mirror anastigmat design that minimizes astigmatism, coma, and spherical aberration. (Chromatic aberration is not a problem in reflectors because dispersion only occurs when light is refracted.) A two-mirror design couldn’t be used in such a wide-field telescope without severe image quality issues, at least comparatively speaking.

https://en.wikipedia.org/wiki/Three-mirror_anastigmat

I'm not asking this skeptically, but out of curiosity and ignorance- how the heck does a 3 mirror setup make sense? Can't a single mirror always accomplish the same goal? (for that matter, why mirrors, with their blind spot where the sensor is, instead of lenses?)

> how the heck does a 3 mirror setup make sense? Can't a single mirror always accomplish the same goal?

Easier to make three smaller mirrors than one big one.

> why mirrors, with their blind spot where the sensor is, instead of lenses?

"Mirrors don't cause chromatic aberration and they are easier and cheaper to build large. They are also easier to mount because the back of the mirror can be used to attach to the mount" [1]. (And there are off-axis designs that avoid obstructing the incoming light.)

[1] https://lco.global/spacebook/telescopes/reflecting-telescope...

That's not intuitive to me, but I recognize that the problem is on my end. Thanks for explaining!

The image from a single curved mirror or lens will have distortions that can be reduced by adding additional curved mirrors/lenses. It's also harder to make and use large diameter lenses than mirrors. Each lens has two surfaces that need to be aligned without the possibility of post-fabrication calibration. Weight and sagging are a bigger problem for lenses. Mirrors can be thinner, partially hollow, and can have mechanical support behind them without blocking light. There are further considerations that might favor mirrors, like material cost and reaction to temperature changes. If nothing else, bending the light path back and forth with mirrors means the telescope can be shorter, easier to point, and will fit in a smaller building.

The largest exclusively lens based ("refractor") telescopes got up to about 1 meter diameter before the trade offs caused a shift to mirrors for larger apertures. Even so, it's common to have lenses near the focal plane of a mirror based ("reflector") telescope to improve the image. Vera Rubin is like that, including a 1.5 meter lens (among others) near the sensor.

The sensor doesn't actually form a blind spot in the image, because it is severely out of focus. Obstructions do affect the pattern of light a star forms on the sensor, but it's all relative, and no mirror or lens can produce perfect images.

Thanks for explaining!

Watching all that monster-size equipment makes me think of the Olympics: is a lot of money being poured into a single use ?

Was this equipment pre-existing or purpose-built ?

If the latter, will any get re-used/repurposed for other endeavors ?

There are a bunch of Vera Rubin-sized mirrors in orbit, that's why this one is the size that it is.

No there’s not -JWST is largest mirror in orbit and it is a few meters smaller, despite being many smaller mirrors. The mirror size is largely due to a tunnel size on the way to the summit in Chile, and applies to the other nearby telescopes.

It’s also really hard to ship mirrors much larger than this on a boat.

Apologies, I was thinking Roman not Rubin.

as much as I'd like to read this, the mobile website is hot garbage due to the number of ads.. if anyone has a better source, I'm all ears

https://www.firefox.com https://github.com/dhowe/AdNauseam

Use Brave if you're on iOS. Firefox on iOS doesn't block ads IIRC

no adnauseam on ff on iOS

Via works great: https://viayoo.com/en/

Reply on news.ycombinator.com