1 00:00:00,000 --> 00:00:05,400 As a cosmic photographer, NASA's Hubble  Space Telescope has taken over a million   2 00:00:05,400 --> 00:00:10,680 snapshots documenting the universe.  These images illustrate, explain,   3 00:00:10,680 --> 00:00:16,920 and inspire us with their grandeur, but may  not match what we'd see with our own eyes.   4 00:00:16,920 --> 00:00:21,540 That's because Hubble sees  light beyond our sensitivity. 5 00:00:21,540 --> 00:00:25,980 Our eyes only sense a small  fraction of the universe's light. 6 00:00:25,980 --> 00:00:31,765 This tiny band of wavelengths, called the  visible spectrum, holds every color in the rainbow. 7 00:00:31,765 --> 00:00:36,900 Light outside that span, with longer or  shorter wavelengths, is invisible to our eyes. 8 00:00:36,900 --> 00:00:42,480 But those invisible wavelengths can  tell us so much more about the universe. 9 00:00:42,480 --> 00:00:47,700 Hubble houses six scientific instruments that  observe at different wavelengths. Together,   10 00:00:47,700 --> 00:00:52,500 they expand our vision into  infrared and ultraviolet light. 11 00:00:52,500 --> 00:00:57,180 That doesn't mean Hubble can show us  never-before-seen colors. In fact,   12 00:00:57,180 --> 00:01:01,860 the telescope can only see the  universe in shades of gray. 13 00:01:01,860 --> 00:01:04,320 Seeing in black and white allows Hubble to detect   14 00:01:04,320 --> 00:01:09,300 subtle differences in the light’s intensity.  If one wavelength is brighter than another,   15 00:01:09,300 --> 00:01:13,685 that tells us something about  the science of that object.   16 00:01:13,685 --> 00:01:19,260 But, because color helps humans interpret what  we see, NASA specialists work to process and   17 00:01:19,260 --> 00:01:23,280 colorize publicly available Hubble  data into more accessible images.  18 00:01:23,280 --> 00:01:28,020 When Hubble snaps a photo, it puts  a filter in front of its detector,   19 00:01:28,020 --> 00:01:31,080 allowing specific wavelengths to pass through. 20 00:01:31,080 --> 00:01:34,620 Broadband filters let in a wide range of light. 21 00:01:34,620 --> 00:01:37,020 Narrowband filters are more selective,   22 00:01:37,020 --> 00:01:42,840 isolating light from individual elements  like hydrogen, oxygen, and sulfur. 23 00:01:42,840 --> 00:01:48,780 Hubble observes the same object multiple times  using different filters. Image processors then   24 00:01:48,780 --> 00:01:52,860 assign those images a color based  on their filtered wavelength. The   25 00:01:52,860 --> 00:01:57,000 longest wavelength becomes red, medium  becomes green, and the shortest blue,   26 00:01:57,000 --> 00:02:02,820 corresponding to the light sensors in our  eyes. Combining them gives us a color image,   27 00:02:02,820 --> 00:02:07,800 showcasing characteristics we  can't make out in black and white. 28 00:02:07,800 --> 00:02:13,860 Adding color reveals the underlying science  in every image. It’s like translating words   29 00:02:13,860 --> 00:02:19,860 into another language, making sure no  information is lost. Some words have an   30 00:02:19,860 --> 00:02:24,900 exact counterpart—the meaning remains the same  when you swap them. Hubble's true-color photos   31 00:02:24,900 --> 00:02:32,040 are like that. They are a direct translation,  using broad filters in wavelengths we can see. 32 00:02:32,040 --> 00:02:35,400 Other words can't be translated directly. When we   33 00:02:35,400 --> 00:02:38,880 use narrowband filters or peer  outside the visible spectrum,   34 00:02:38,880 --> 00:02:46,860 it's like translating words with no one-word  replacement. Easily done, but requires more work. 35 00:02:46,860 --> 00:02:53,580 Narrowband images highlight the concentration of  important elements. Infrared images are like heat   36 00:02:53,580 --> 00:03:01,707 maps, helping us spot newborn stars in dark, dusty  clouds and peer further back in time and space.   37 00:03:01,707 --> 00:03:09,360 In ultraviolet, we uncover active aurorae on Jupiter  and learn how young, massive stars develop. 38 00:03:09,360 --> 00:03:13,740 Image processors also clean up  artifacts—signatures in an image   39 00:03:13,740 --> 00:03:19,500 that aren't produced by the observed target.   As sensors age, some pixels become imperfect,   40 00:03:19,500 --> 00:03:26,400 returning too much electrical charge, or  not enough. Artifacts can leave behind odd   41 00:03:26,400 --> 00:03:33,780 shapes or return images without any true black.  These effects can be calibrated and removed.  42 00:03:33,780 --> 00:03:39,660 Other artifacts come from the dynamic environment  of space. Even the best photographers get   43 00:03:39,660 --> 00:03:45,660 photobombed; in Hubble's case, the culprits  are asteroids, spacecraft or debris trails,   44 00:03:45,660 --> 00:03:52,140 and high-energy particles called cosmic rays.  By combining and aligning multiple observations,   45 00:03:52,140 --> 00:03:58,380 image processors can identify them and  piece together an artifact-free image. 46 00:03:58,380 --> 00:04:03,060 Without processing, many Hubble images  would be divided down the middle.   47 00:04:03,060 --> 00:04:09,600 This line, called a chip gap, is the tiny space  between some camera sensors. Hubble moves slightly   48 00:04:09,600 --> 00:04:18,534 with each observation, allowing image processors  to fill the gap and replace faulty pixels. 49 00:04:18,534 --> 00:04:21,780 And because there's no  natural up or down in space,   50 00:04:21,780 --> 00:04:27,060 processors decide how to  rotate and frame the image. 51 00:04:27,060 --> 00:04:31,380 It's a time consuming procedure.  Simple images take about a week,   52 00:04:31,380 --> 00:04:39,660 while large mosaics, stitched together from  many observations, can take a month to process. 53 00:04:39,660 --> 00:04:45,240 Hubble images may not be what we'd see  firsthand. Instead, they are tools for   54 00:04:45,240 --> 00:04:51,959 understanding science at a glance, shedding light  on otherwise invisible views of our universe. 55 00:04:51,959 --> 00:04:56,366 Follow us on social media @NASAHubble