9
$\begingroup$

At the end of December 2015, a giant "painting" made of 4 million color-sorted recycled bottles was put on display in Keelung, Taiwan. The Indian Express says it's 53 hectares, which is 530,000 square meters, so it's probably roughly 0.5 x 1 km in size. At 400km, that makes it roughly a quarter of the diameter of the moon visually - if I've done my arithmetic correctly.

As a reference, the 3U Cubesat PlanetLabs Doves have a resolution of 3 to 5 meters.

If they knew about it, could an astronaut on the ISS have been able to photograph some artificial, artistic pattern similar to this either with a hand-held camera (with a long lens) or with the downward-pointing telescope on the station?

I understand that either the hand-held camera or the telescope might need to be moving to compensate for the relative motion between the ISS and the surface. Besides various forms of architecture and civil engineering, has any art on earth been photographed from the ISS?

Image from Indian Express

enter image description here

$\endgroup$
  • 2
    $\begingroup$ For reference: various capital city photos, from the ISS, are available here. I think resolution of the order of 10-20m seems reasonable from the better examples of these. Just a guide to what's possible... $\endgroup$ – Andy Jun 20 '16 at 16:27
10
$\begingroup$

I like @PearsonArtPhoto's answer, though the considerations about exposure time may not be correct. I was looking for something more visual.

The website states an area of 0,53m² of the Van Gogh image, and since the original is 92.1 × 73.7 cm², the large version must be about 0.817km wide.

During his year in space, Scott Kelly took tons of photos, one of New York from straight above at night. And Central Park is about 4.07 × 0,86 km² in size, which means its small side is roughly as long as the long side of the Van Gogh.

So here is the picture from Manhattan with the Van Gogh above Central Park. The Van Gogh was scaled by a factor derived from the size of the original, of Central Park in reality, and the pixel size of Central Park on the photo, so it's quite exact. (Width: 105px) I blurred the right one until I felt that its sharpness is about the same as that of the entire image.

Finally, I cropped the picture to the interesting area, but didn't scale it.

enter image description here
(Source of original: https://www.nasa.gov/content/best-year-in-space-photographs-from-scott-kelly)

What's quite interesting now is the EXIF data of the photo:

  • It was taken with a Nikon D4, making photos of 4.928 × 3.280 px² (16.4MP) with its full frame (36mm) sensor. The size of the original photo is the same, so it was not cropped or scaled.
  • The lens was a fixed-length 400mm f/2.8, and the photo was taken at f/2.8 to collect as much light as possible. Note: A higher aperture of may be f/5.6 collects less light, but would give a sharper image. (Though this effect should be low for this lens).
    By the way: Assuming an altitude of 400km, the width of the Van Gogh should be 112px, which fits my 105px quite well. So, no teleconverter was used!
  • Exposure time was 1/8s, during which the ISS moves about 1km. But there's absolutely no in-motion unsharpness in the Photo, so the camera tracked the motive very well.
    I'm wondering, because usually it's very difficult to shoot a sharp 1/8s photo at 400mm when holding the camera in your shaking hands. Such lenses have image stabilization, but this doesn't work nicely when you have to track the motive. May be, zero gravity and the weight of the lens helps a lot. Finally, the photo was processed with Photoshop, and some unsharpness could have been removed.

So, the Van Gogh is about 100px wide when photographed with a 400mm lens on a 16.4MP full frame camera. If the ISS has a 800mm and a converter to get 1120mm, it would be 200px or 280px wide. If a camera with APS-C sensor is used, the width increases to 448px, and if this has 24MP instead of 16MP, the width would be 546px.

However, sharpness of a lens is limited. You can't increase sharpness by using smaller and smaller pixels. While I think you would get a sharper image with the 800mm plus teleconverter, you would not gain more sharpness with an other camera.

Conclusion:

Playing with the numbers, the Van Gogh would still be small compared to the entire photo size, but large enough to be noticed. Sharpness is limited, but looking at the level of detail of the photo, the Van Gogh should be recognizable even when photographed with a 400mm lens.

$\endgroup$
  • $\begingroup$ yep - as you mentioned, "hand-held" cameras don't have to be! :-) I don't know what image stabilization from a lens means, but he could certainly take many frames and de-blurr by overlaying them by shifting, but then you'd loose some of the edge, and you said the images have all the pixels that the metadata says. Hmm... $\endgroup$ – uhoh Jun 22 '16 at 10:50
  • $\begingroup$ @uhoh: I.S. means the lens measures the shatter and moved some lens elements laterally to compensate it. Manufacturers say you can increase exposure by a factor of 22, so a 1/8s photo with I.S. is as sharp as a 6/100s photo without. The factor usually isn't that large (may be for this lense), and I.S. usually doesn't work if you follow the motive. Multiple frames don't make sense here, too, because for this, the ISS is really too fast. $\endgroup$ – sweber Jun 22 '16 at 12:05
  • $\begingroup$ OK got it! Thanks for the really thorough answer and "simulation" too! $\endgroup$ – uhoh Jun 22 '16 at 12:43
  • $\begingroup$ Oh, of course 1km/s! $\endgroup$ – sweber Jun 22 '16 at 13:45
  • 1
    $\begingroup$ A few things. First of all, that assumes a high contrast as shot from above. From what I can tell, the image wouldn't have that characteristic. Secondly, NYC is photographed at night. A much higher exposure time is required for nighttime shooting, as you noted. The blurring does allow for it to be recognized still, so I'll admit, my numbers were pessimistic. Still, nice work. $\endgroup$ – PearsonArtPhoto Jun 22 '16 at 14:03
6
$\begingroup$

The ISS, at it's closest, is, say, 500 km (Can be a bit closer, but let's not stretch it). That would require a FOV of 1/500, or about 0.1 degrees. That would require a camera with a lens of about 20000 mm to get a full view, given a standard sensor. That isn't going to happen! But still, a low resolution view of the image is entirely possible. The ISS's best telephoto lens is an 800 mm lens with a teleconverter to 1120mm, and with a crop factor, it's FOV is about 1.2 degrees apart. So if you have a, say, 20 megapixel image, about 200,000 pixels will cover the image, which is manageable.

Okay, so how fast would such an image need to be taken to minimize distortion? Let's try to limit distortion to 10m, which is probably a bit high for such an image. The space station moves at 7667 m/s. So you'd have to get the image in around 1/1500th of a second. Most cameras have something called a shutter sync. I've checked and some of the ISS cameras are at 1/8000th of a second, so a high shutter speed image could be taken that would image it.

To my knowledge, no such thing has been photographed from the ISS. They simply don't have the resolution to really do it justice. But there are several art worthy photos taken, if you look around, of places on Earth.

$\endgroup$
  • $\begingroup$ You don't need to zoom in so far with modern cameras. You can snap a photo of the moon with a 50mm lens, crop, and still recognize that it's the moon. I should edit the question to say that it doesn't have to be a beautiful photo - a quad-copter can do that. Even if its fuzzy, it's still shot by astronauts from space. $\endgroup$ – uhoh Jun 20 '16 at 15:55
  • 1
    $\begingroup$ The Moon is larger, and higher contrast. It could be done with less, but the Issue could do it justice, with the best equipment they have. $\endgroup$ – PearsonArtPhoto Jun 20 '16 at 15:57
  • $\begingroup$ Wow, OK I see what you mean. Do you mean there were photos of earth that fast, or just that there were camera bodies that could shoot that fast? I assume the earth can be quite bright, but that's really fast. If the high speed means under exposure, can they shoot say a burst of 10 or 20 and then shift them all to make one higher exposure image? $\endgroup$ – uhoh Jun 20 '16 at 16:02
  • 1
    $\begingroup$ The cameras are capable, with high ISO is required, to max the shutter speed. $\endgroup$ – PearsonArtPhoto Jun 20 '16 at 16:09
  • $\begingroup$ So actually this "painting" wasn't designed for space viewing. If most of the color or intensity contrast is 10 meters or below, it may wash out and just look gray. You might want 10-20 meter wide bands. By the way, would the streaking or blurring due to motion just be in one direction? Perpendicular to the motion, the resolution would still be limited by everything else, but not by the blurring. If so, that might be a consideration. $\endgroup$ – uhoh Jun 20 '16 at 17:13
4
$\begingroup$

Loosely on the theme of art photographed from space:

A set of temporary floating piers have recently been set up on Lake Iseo, Italy as a piece of interactive art.

An animated picture from the Sentinel 2A satellite has just been released. This shows the bright yellow piers (and a few stages of construction). The piers are 16 metres wide and covered with bright yellow fabric.

Floating piers on lake Iseo, by Sentinal-2A

Note - this isn't from the ISS though. A version taken by hand might not match the resolution shown here.

$\endgroup$
  • 1
    $\begingroup$ OK - this is definitely art photographed from space - the last part of my question has been answered - thanks! I have a hunch there should be plenty more of this artists work on satellite photographs, considering it's often quite large in scale and outdoors. We can debate (elsewhere please) if it is space-art, or simply art photographed from space. You can never really be sure with art ;-) $\endgroup$ – uhoh Jun 22 '16 at 16:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.