Low-Cost Approach to Scanning Historic Glass Plates Yields an Astronomical Surprise. Technical details at Precise Photometric Measurements from a 1903 Photographic Plate Using a Commercial Scanner.

Professional astrophotography used to be done on emulsion-coasted glass places. That was how astromical discoveries were made for nearly a century.

More than an estimated 2.4 million glass plates are out there in collections in North America alone. These were taken starting in the 1890s right up until the 1970s, when CCD (Charged Couple Device) detectors started to come online for astronomy. Of these, only an estimated 400,000 plates have been digitized to research quality

The team in this article has found a much cheaper way to proceed with this process, using off-the-shelf hardware.

Their efforts were immediately rewarded when a 1903 plate showed a “guest star,” i.e. a supernova candidate, in the galaxy NCC7331 that did not appear on the modern sky survey that they were using to check their results. How do you find a single such star on a plate lots of stars?

“We were going through the image of the galaxy on the plate as part of our analysis, which involved comparing the plate with a modern image of the same field of sky. At one point, we blinked (rapidly alternated) between the two pictures, and noticed what appeared to be a star present on the plate image.”

If you have two otherwise identical pictures, an object on one but not the other will appear to blink if you superimpose the two images, then rapidly switch back and forth between the two. The trick is to get the “otherwise identical pictures.” This team had the benefit of modern computer techonology, but the technique has been around for decades. Clyde Tombaugh used a specialized device to discover Pluto in 1930:

Tombaugh operated the instrument, built by the Zeiss company and popularly called a “blink comparator,” by rotating a small dial that flipped a mirror back and forth between the beams from two microscopes. Blinking between two plates taken on January 23 and 29, 1930, he found something. The below animation is flipping between the two plates…. It’s not an easy job, to be sure. The time exposures do not always yield the same brightness, and the orientation of the two fields is not exact

Getting the plates to match well enough for this to work, and analyzing the results was an incredibly finicky process. Fred Hoyle described it in his 1957 novel The Black Cloud:

So on this late January morning of 8th January, 1964, Jensen was down in the basement of the Observatory buildings setting up an instrument known as the ‘blinker’ . As its name implies, the ‘blinker’ was a device that enabled him to look first at one plate, then at the other, then back to the first one again, and so on in fairly rapid succession. When this was done, any star that had changed appreciably during the time interval between the taking of the two plates stood out as an oscillating or ‘blinking’ point of light , while on the other hand the vast majority of stars that had not changed remained quite steady . In this way it was possible to pick out with comparative ease the one star in ten thousand or so that had changed. Enormous labour was therefore saved because every single star did not have to be examined separately.

Great care was needed in preparing plates for use in the ‘blinker’ .They must not only be taken with the same instrument, but so far as possible must be shot under identical conditions. They must have the same exposure times and their development must be as similar as the observing astronomer can contrive . This explains why Jensen had been so careful about his exposures and development.

His difficulty now was that exploding stars are not the only sort to show changes. Although the great majority of stars do not change, there are a number of brands of oscillating stars, all of which ‘blink’ in the manner just described . Such ordinary oscillators had to be checked separately and eliminated from the search . Jensen had estimated that he would probably have to check and eliminate the est part of ten thousand ordinary oscillators before he found one supernova . Mostly he would reject a ‘blinker’ after a short examination, but sometimes there were doubtful cases. Then he would have to resort to a star catalogue, and this meant measuring up the exact position of the star in question. So all in all there was quite a bit of work to do before he got through his pile of plates—work that was not a little tedious.