The World's Largest Digital Camera

[CT_Shooter]

The Vera C. Rubin Observatory has the world's largest digital camera (three billion pixels). It's an amazing scientific and technical achievement that took its first images in April 2025 and we'll see some of them next Monday.

I've printed the NY Times article to a PDF file and attached it for those who might be curious. viewtopic.php?f=19&t=16784

Every 30 seconds, the observatory will point to a different part of the sky, capturing an area greater than 40 full moons. The end result will be a catalog of 20 billion galaxies and 20 billion stars in six different colors, stretched across not only the wide expanse of space, but also of time.

Vera Rubin Observatory Brings Universe’s Darkest Mysteries Into Focus - The New York Times.pdf

(30.34 MiB) Downloaded 96 times

[Ernie]

I am guessing that you would have to look at sections at a time. Not at all sure how large 3 billion pixels is.

[Mags]

I am guessing that you would have to look at sections at a time. Not at all sure how large 3 billion pixels is.

The pixel array is about 25 x 25 inches. It's made up of 189 smaller 16mega pixel sensors, 1.6 inches square. Each pixel is about 10 microns (0.000393701 in.) in size.
To display one image full size, it would take about 378 4K ultra-high-definition TV screens.

https://www.inverse.com/innovation/larg ... ion-pixels

[Ernie]

Thanks Mags.

[clovishound]

Just to put things in perspective, the most common commercially available digital cameras today are about 24 megapixels. That's 24 million pixels. The higher resolution cameras are in the 45 - 50 megapixel range. The ultra high resolution (that's pronounced: "VERY EXPENSIVE!") cameras are 100 megapixels.

It would take 60 images with my camera to equal the resolution this camera produces.

[CT_Shooter]

I am guessing that you would have to look at sections at a time. Not at all sure how large 3 billion pixels is.

The pixel array is about 25 x 25 inches. It's made up of 189 smaller 16mega pixel sensors, 1.6 inches square. Each pixel is about 10 microns (0.000393701 in.) in size.
To display one image full size, it would take about 378 4K ultra-high-definition TV screens.

https://www.inverse.com/innovation/larg ... ion-pixels

Yes. Thanks, Mags. It is a mind boggling amount of data that are being collected by this system. I'm sure that "looking" at it will take lots of different forms. We should get some sort of slideshow next week. But, it has a long-term goal to "chart the depths of the entire southern sky every three nights for the next 10 years, resulting in a crisp motion picture of the ever-changing cosmos above". Scientifically, they're hoping to unlock the mystery of dark matter and dark energy. Personally, as a layman and consumer, I can imagine an awesome VR experience (maybe at Universal theme park. )

[Mags]

... I can imagine an awesome VR experience (maybe at Universal theme park. )

Maybe they'll have a website with images we can zoom in on forever. That would be awesome!

[rickhem]

I am guessing that you would have to look at sections at a time. Not at all sure how large 3 billion pixels is.

The pixel array is about 25 x 25 inches. It's made up of 189 smaller 16mega pixel sensors, 1.6 inches square. Each pixel is about 10 microns (0.000393701 in.) in size.
To display one image full size, it would take about 378 4K ultra-high-definition TV screens.

https://www.inverse.com/innovation/larg ... ion-pixels

So for me to appreciate the scale of this, I took the 3 billion pixel stat for this camera, and calculated the overall size of the photograph it would produce. Using your stat for 1 pixel being equal to 10 microns, that gives us 30 billion pixels from this camera. So how big is 30 billion pixels? I calculated it in miles as that is something we can get our minds to comprehend a little easier.

30 billion pixels equates to 18.64113576712 miles.

That is one heck of an image to study.

[Mags]

Today is the day for the first pictures...
https://rubinobservatory.org/

[clovishound]

I'm a little surprised at the sensors they used to make this array. A modern high resolution sensor available in many modern digital cameras is 45-50 megapixels for a 1.4 x 1 inch sensor. Of course larger pixel sizes IE a less dense sensor, generally has better low light performance. Since they are dealing with extremely low light levels, albeit, at long, multiple exposures, it probably shouldn't surprise me.

Of course, megapixels isn't everything. A lot of factors determine how large you can blow up an image and still have good apparent sharpness and detail. You can blow up a 12 megapixel image to billboard size, and it still looks good....... from 1/2 a mile down the road. Don't get me wrong, I would really hate to "back up" to a 24 MP sensor camera, even though I realize the extra resolution is not needed for most images and uses. With something like a telescope they can do things like multiple exposures, pixel shifting etc to get the most from their gear.

Still, this is an impressive piece of gear.

[clovishound]

So for me to appreciate the scale of this, I took the 3 billion pixel stat for this camera, and calculated the overall size of the photograph it would produce. Using your stat for 1 pixel being equal to 10 microns, that gives us 30 billion pixels from this camera. So how big is 30 billion pixels? I calculated it in miles as that is something we can get our minds to comprehend a little easier.

30 billion pixels equates to 18.64113576712 miles.

That is one heck of an image to study.

While 30 billion pixels is very impressive, I think you are looking at an image size 1 pixel on one side, and 30 billion on the other, rather than a square image. If my math is correct, it will yield an image a tad over 21 inches square. Keep in mind that this is just the image size from the sensor. That image is normally enlarged to a convenient viewing size. A measly 24 MP image is 1 in by 1.4 inches and can easily be enlarged to 16 x 20 inches and viewed from a foot away and still be perceived to be sharp and high resolution.

Check my math: 30000000000 pixels x 10 microns find the square root for area = 547,722 microns convert to inches = 21.56 inches.

[Mags]

I'm a little surprised at the sensors they used to make this array. A modern high resolution sensor available in many modern digital cameras is 45-50 megapixels for a 1.4 x 1 inch sensor. Of course larger pixel sizes IE a less dense sensor, generally has better low light performance. Since they are dealing with extremely low light levels, albeit, at long, multiple exposures, it probably shouldn't surprise me.

Of course, megapixels isn't everything. A lot of factors determine how large you can blow up an image and still have good apparent sharpness and detail. You can blow up a 12 megapixel image to billboard size, and it still looks good....... from 1/2 a mile down the road. Don't get me wrong, I would really hate to "back up" to a 24 MP sensor camera, even though I realize the extra resolution is not needed for most images and uses. With something like a telescope they can do things like multiple exposures, pixel shifting etc to get the most from their gear.

Still, this is an impressive piece of gear.

Details of the array are also in an earlier post.
https://henryrifleforums.com/viewtopic. ... 19#p256243

[clovishound]

Thanks, I forgot the total sensor size was included in the original post. I calculated the size based on the number of pixels and approximate pixel size. It was close, but I guess round off, and other issues take their toll.

[Mags]

Thanks, I forgot the total sensor size was included in the original post. I calculated the size based on the number of pixels and approximate pixel size. It was close, but I guess round off, and other issues take their toll.

I think the difference, between 21 and 25 inches, is accounted for mostly in the gaps between the matrix of 189 sensor panels.

[clovishound]

Yeah, like a 2.0 liter engine is actually 1.976 liters.

The other thing is that the calculation I made was for a complete square. It looks like they didn't put sensors in the corners. Most digital cameras use a single square/rectangular sensor and the lens' circular image has part of the image falling beyond the sensor's edges. Since they were using multiple sensors, they were able to roughly follow the circular image and not waste sensors that had no viable image falling on them.