Between 18 and 28 December 1995, the Hubble Space Telescope repeatedly photographed one deliberately unpromising patch of sky near the handle of the Big Dipper. The Hubble Deep Field combined 342 exposures gathered over 10 days and revealed roughly 3,000 galaxies.

NASA describes the field as about the apparent size of a pinhead held at arm’s length. Some of the faintest recorded light had been travelling for about 10 billion years, showing galaxies as they appeared when the Universe was much younger.

The field was not empty in a physical sense. It was selected to contain as little bright foreground clutter as possible, allowing a long exposure to recover sources that shorter observations left below the detection threshold.

The field was chosen to look ordinary

Robert Williams, then director of the Space Telescope Science Institute, committed part of his director’s discretionary observing time to the project. NASA’s history of the Deep Field says a team spent nearly a year planning the observation and taking test exposures.

The target had to lie away from the bright plane of the Milky Way. It could not contain a prominent nearby star, galaxy or galaxy cluster whose light would overwhelm fainter sources. It also needed to fall within a region Hubble could view for extended periods without interference from the Earth, Sun or Moon.

The team selected a high-latitude field in Ursa Major. Calling it empty meant that catalogues and test images showed little obvious foreground structure. It did not mean astronomers had established that the line of sight contained no distant galaxies.

A long exposure does not create objects; it collects enough photons to make faint ones detectable.

Ten days meant 342 separate exposures

Hubble did not hold one shutter open continuously for 10 days. The telescope orbited Earth about 150 times during the programme, collecting separate frames whenever the target was visible and conditions allowed. ESA’s Hubble summary puts the useful exposure time at more than 100 hours.

The Wide Field and Planetary Camera 2, or WFPC2, recorded the field through four filters spanning ultraviolet, blue, red and near-infrared wavelengths. Combining those passbands produced the colour view and supplied information about each source’s spectrum. Repeated frames also helped the processing team distinguish detector defects and energetic-particle strikes from real objects.

The pinhead description is an apparent-size analogy. A NASA archive page for the original mosaic uses that wording, while the 1996 release compared the field with a dime seen from about 75 feet away.

Either comparison describes a very small sample of the sky.

The galaxy count changed after the first release

NASA’s January 1996 announcement referred to at least 1,500 galaxies. Its current archive says later work raised the estimate to around 3,000. The title’s figure is a later count, not the number used when the image was first released.

Counts at this depth depend partly on the detection threshold and on how overlapping or irregular patches of light are separated. Some sources occupy only a small number of pixels. “Roughly 3,000” is the appropriate level of precision.

Distance also needs careful wording. The image supplies brightness and colour measurements; redshift estimates and follow-up observations provide distance information. NASA says light from some of the farthest galaxies travelled for about 10 billion years. That is a lookback time, not a simple statement of their present-day distance in an expanding Universe.

The image also combines nearby and remote objects, so it does not show one moment 10 billion years ago. Different galaxies appear at different stages of the past.

Astronauts installed the camera two years earlier

The Deep Field has a direct place in human-spaceflight history. Hubble reached orbit aboard space shuttle Discovery in April 1990, but spherical aberration in its primary mirror limited the sharpness and sensitivity expected from the telescope.

Seven astronauts flew Endeavour on STS-61 in December 1993 for Hubble’s first servicing mission. Across five spacewalks, the crew replaced degraded equipment, installed corrective optics for several instruments and exchanged the original Wide Field and Planetary Camera for WFPC2.

WFPC2 carried its own internal optical correction. According to NASA’s STS-61 history, Story Musgrave and Jeffrey Hoffman completed the camera exchange during the mission’s third spacewalk. Two years later, that astronaut-installed instrument recorded every exposure in the Deep Field.

The image was produced by an automated observatory and processed on the ground. The crewed contribution was specific: Hubble’s serviceable design gave astronauts access to restore its intended optical performance and install the camera used for the observation.

The image was a public dataset, not a final answer

Williams and the Hubble team released the reduced images without the usual proprietary period. Their 1996 paper, “The Hubble Deep Field: Observations, Data Reduction, and Galaxy Photometry” in The Astronomical Journal, documented the field selection, observing plan, processing and source catalogue.

One narrow field could not establish whether every direction contained the same mixture of galaxies. That limitation helped motivate the Hubble Deep Field South in 1998. Later programmes, including the Hubble Ultra Deep Field, used newer cameras and longer accumulated exposures to detect still fainter populations.

The original Deep Field did not settle galaxy formation in one image. It provided a documented set of observations that independent groups could examine, compare and challenge.

Its connection to human spaceflight remains equally concrete: the 1995 dataset depended on replaceable orbital hardware and on the 1993 crew that installed the corrected camera.