Illuminated Universe

Hubble Heritage Archive

Hubble Heritage Archive: Images and Gallery of the Project, 1998-2016

Eighteen years of pictures made from Hubble science data, how a small team turned gray exposures into color, and every object in the series.

The Whirlpool Galaxy M51, a face-on spiral with pink star-forming knots along its arms, connected to a smaller yellowish companion galaxy
The Whirlpool Galaxy (M51) and its companion, photographed with the Advanced Camera for Surveys in January 2005 and released that April. Image: NASA, ESA, S. Beckwith (STScI), and The Hubble Heritage Team (STScI/AURA)

Hubble Heritage images are the pictures most people have in mind when they think of the Hubble Space Telescope: the Whirlpool Galaxy, the Sombrero, the Cat's Eye, the Ring Nebula. Between October 1998 and 2016 a small group at the Space Telescope Science Institute in Baltimore released one new image roughly every month, nearly all of them made from data that had been collected for research. This page explains what the project was, how a Heritage picture was put together, and links to every object in the series.

The Hubble Heritage Project was based at the Space Telescope Science Institute (STScI), which the Association of Universities for Research in Astronomy (AURA) operates for NASA. The team was deliberately mixed. Some members were professional astronomers, others came from image processing and design, and decisions about framing, color and contrast were made jointly. Keith Noll led the project as principal investigator, and in 2007 he received the American Astronomical Society's Education Prize for creating and running it.

What the Hubble Heritage archive holds

The raw material came from two places. Most images were mined from the public Hubble data archive, which by the late 1990s already held close to a decade of exposures. Researchers had taken those exposures to answer scientific questions, so the visual result was a side effect. That created gaps. Some of the most photogenic objects in the sky had never been proposed as research targets, and favorites that had been observed often lacked exposures in enough filters for a color picture, or covered only a small, unrecognizable piece of the object.

To fill those gaps the project was given a small allocation of observing time, drawn from the STScI director's discretionary time. Those orbits had two jobs: to produce a picture and to produce data that astronomers could still analyze. The Heritage data went into the public Hubble archive at the same time as each picture was released. A few outside astronomers joined specific observations as guest collaborators, and on at least one occasion the public chose the target. For the polar-ring galaxy NGC 4650A, about 8,000 internet voters picked the object from a list of candidates before the team scheduled it.

The first new observation the team made itself was the Ring Nebula in Lyra, imaged with the Wide Field and Planetary Camera 2 on October 16, 1998 and released on January 6, 1999. Over the following years the Hubble Heritage archive grew to cover planets, star clusters, nebulae, dying stars and distant galaxies. The essay on the aims of the Hubble Heritage Project looks at what the series changed about public astronomy pictures.

How Heritage images were composed

Every Hubble camera records only brightness. A detector counts how much light lands on each pixel and produces a black-and-white frame. Color enters through filters placed in front of the detector, each passing a chosen band of wavelengths. Taking several exposures through different filters and assigning each one a hue is how all Hubble Heritage images were built.

When three broadband filters roughly matched the response of the human eye, red, green and blue could be assigned in the obvious way and the result was close to natural color. The Sombrero Galaxy picture of 2003 was made like this. Many research programs, though, used narrowband filters that isolate the light of one element in one state, such as hydrogen-alpha, doubly ionized oxygen or ionized sulfur. Those exposures cannot add up to natural color, because the eye would see most of that light as nearly the same shade of red.

The team handled this in two ways. A chromatic ordering keeps the sequence of wavelengths: the shortest filter is shown blue, the longest red, and the middle one green. A composite ordering breaks that sequence when doing so separates structures that would otherwise blur together. The project's own notes stated that showing subtle structure and the physical processes in a nebula generally took priority over natural color. The page Are the Colors Real? covers the reasoning behind these choices.

Contrast was handled with methods that astronomers use to inspect data and with the digital equivalents of darkroom dodging and burning. The rule the team set itself was that no feature could appear in the final image unless it was present in the data.

The challenges behind each set of images

Hubble does not point wherever it is asked at short notice. A research observation starts as a proposal submitted in response to an STScI call for proposals, reviewed by a panel of scientists for merit and technical feasibility. An accepted program is assigned a Program Coordinator, responsible for scheduling, and a Contact Scientist, responsible for the scientific details.

Scheduling has hard limits. The telescope cannot point too close to the Sun or the Moon, whose light could damage the instruments. It needs guide stars to lock onto during an exposure, and a field with too few stars, or a crowded field with too many, can make that difficult. Accepted observations go onto a week-long flight calendar, and those calendars are built about three weeks in advance.

Framing was a problem of its own with the Wide Field and Planetary Camera 2, the main camera of the early Heritage years. It had four detectors, and one of them had a finer pixel scale than the other three. When all four frames are scaled to match, the combined field has a stepped, chevron outline. Observers had to choose where to place the target and how to rotate the telescope so the whole object fell inside that shape. NGC 4650A, for example, was deliberately placed lengthwise across two of the four chips, just above the seam between them.

The NGC 4650A observation of April 1999 shows each stage. The raw WFPC2 frames, straight from the telescope, are speckled with dots and short streaks. These are cosmic rays, energetic charged particles that struck the detector during the exposure. Taking several exposures through the same filter and comparing them lets the software reject anything that appears in only one frame, so the combined image is clean and the background galaxies and foreground stars stand out.

Placing the single-filter images side by side at similar display levels shows how the galaxy changes with wavelength. Stars, the bright center and fine structure in the ring all vary in strength from filter to filter. A first color draft then assigned the reddest filter, F814W, to red, F606W to green and F450W to blue. That chromatic draft already told the main story: a smooth, reddish disk of older stars crossed by a thin ring at nearly a right angle, with dust in the ring reddening and hiding the disk where the two overlap. Several more color renditions were tried before the final version was released on May 6, 1999.

The project also published practical notes on the files themselves. Hubble's cameras of that era produced frames no larger than about 1,600 pixels on a side, so a bigger file did not mean more detail. Enlarging the pixel count beyond the original data only degrades the picture, and for that reason the team never offered wallpaper-sized versions. Heritage images could be reproduced without asking permission when the credit read "NASA, ESA, and The Hubble Heritage Team (STScI/AURA)", while any additional names in a credit line had to be contacted separately.

The series collected recognition from outside astronomy. The Griffith Observatory gave the project a STAR Award in January 1999, three months after its debut. The International Center of Photography in New York presented a 2000 Infinity Award for applied photography. The Ring Nebula picture appeared on a 2000 United States postage stamp honoring Edwin Hubble, Heritage images featured in a 2002 British stamp set, and the Astronomical Society of the Pacific awarded the team its Klumpke-Roberts Award in 2003 for contributions to public understanding of astronomy.

The Hubble Heritage archive pages below each cover one object: what it is, what the data were, which filters were used and how the colors were assigned. They are grouped by type.

Nebulae

Planetary nebulae

Supernova remnants

Galaxies

Stars and clusters

Solar system

Questions readers ask

When did the Hubble Heritage Project run?

The project made its public debut in October 1998 and released its pictures through 2016. In the early years a new image appeared roughly once a month. Later releases were often made jointly with ESA/Hubble, the European outreach office for the telescope, and widely reproduced pictures such as the 2005 Whirlpool Galaxy and the 2016 Bubble Nebula carry the Heritage Team in their credit line.

Were Hubble Heritage images made from science data?

Yes. Most were built from exposures already in the public Hubble archive, taken by research teams for their own studies. When the archive lacked a filter or the full extent of an object, the team used a small amount of its own observing time, and those new data were placed in the archive for astronomers at the same moment the picture was published.

Can Hubble Heritage images be used freely?

The project allowed reproduction without a permission request when the credit line read "NASA, ESA, and The Hubble Heritage Team (STScI/AURA)". Some images carry extra names, such as the principal investigator of the original observations, and in those cases the named people or institutions also had to be asked. The usage terms printed on each image page today are the ones to follow.

Why do some Heritage images have a stepped corner?

Those images come from the Wide Field and Planetary Camera 2. One of its four detectors recorded a smaller patch of sky at higher magnification, and once all four frames are scaled to the same pixel size the combined field has a notch in one corner. The team sometimes cropped it away and sometimes left it in the frame.