For 300 years, we only had a partial view of Rembrandt’s 17th-century masterpiece “The Night Watch.” But now, with the help of Artificial Intelligence (AI), researchers at the Rijksmuseum in the Netherlands were able to reconstruct missing pieces of the painting, giving us a rare view of what it looked like when the Golden Age Dutch artist finished it in 1642.
In 1639, Rembrandt was commissioned to make the painting for a new banqueting hall at the headquarters of the Kloveniers, the civic militia guards (or musketeers) of Amsterdam. The painting was part of a series of seven militia portraits (schuttersstukken), commissioned by Captain Banninck Cocq along with 17 members of his militia. In 1715, “The Night Watch” was moved to what was at the time Amsterdam’s City Hall, now the Royal Palace on Dam Square. Alas, the colossal painting was deemed too large for its new location, so it was trimmed from all four sides, with the largest section being removed from the left side. According to art historians, this was done in order to fit the painting between two columns, a common practice at the time. The trimmed pieces have never been found.
“The fate of the missing pieces of ‘The Night Watch’ remains a great mystery,” said Pieter Roelofs, head of the paintings and sculptures department at the Rijksmuseum in Amsterdam. “Each generation has used the tools available to it to attempt to reconstruct the painting. Now we are doing the same, using the most advanced techniques currently available.”
In 2019, the Rijksmuseum launched “Operation Night Watch” to explore the original composition of the most famous painting in its collection. Based on a copy of the original painting that is believed to have been painted by Dutch artist Gerrit Lundens between 1642 and 1655, the museum’s researchers taught Rembrandt’s technique and use of color to a computer, which used “artificial neural networks” to recreate the missing pieces in Rembrandt’s style.
The recreated panels were printed and temporarily placed around the original painting, which is now on view at the museum and on its website. Conservation work will only begin after researchers finish studying the impact of the added panels on the painting’s composition. However, the museum does not intend to incorporate the lost pieces in the actual restoration of “The Night Watch.” The researchers say that the main benefit of experiencing the original composition is that it “allows for a better comparison with the six other works.”
According to the museum, this is the largest and most wide-ranging research project ever conducted into Rembrandt’s masterpiece.
“This project testifies to the key importance of science and modern techniques in the research being conducted into ‘The Night Watch’,” said Robert Erdmann, a senior scientist at the Rijksmuseum. “It is thanks to artificial intelligence that we can so closely simulate the original painting and the impression it would have made.”
Highlights from the reconstruction reveal two more figures on the painting’s left side: a militiaman and a young boy. The painting’s main figures, Captain Cocq and Lieutenant Willem van Ruytenburch, are now positioned to the right of the center instead of the middle of the canvas. The reconstruction also shows that the “powder boy” in the left foreground is holding onto a balustrade. With more space for movement in front of him, the boy now seems to be running away, ahead of the militia, according to the researchers. On the extreme right side of the canvas, the helmet worn by the militiaman is now complete. There is also more space on the top part of the painting, making the motion of raising the militia’s flag “more convincing,” the researchers say.
“‘The Night Watch’ as it is displayed in the Rijksmuseum is etched into our collective memory,” said Taco Dibbits, director of the Rijksmuseum. “Thanks to this reconstruction, we can now see that the composition as it was painted by Rembrandt was even more dynamic. It is wonderful to be able to now see with our own eyes ‘The Night Watch’ as Rembrandt intended it to be seen.”