I see photoshops offering drum scanning to 1 Giga Pixels and even more. How is a machine from the past able to extract so much information? In addition to how the scanning process functions, does anyone know how lab techs in the 80s 90s handled extremely large amount of data (several GBs) generated by the scanner?
The drum scanner operated by affixing a film or print image to the outside of a cylinder of clear glass. The drum was powered up causing it to spin at high speed. A light source located inside the cylinder outputs a focused spot of light that passes through the glass cylinder and traverses the film.
The color and intensity of this beam of light is thus modified by film. This emerging light then plays on a photo multiplier tube. A photo multiplier tube is old-school, a cousin of the old-fashion radio / TV glass enclosed tubes. Light hitting this tube converted light energy into an electric charge. The photo multiplier was outstanding due to its ability to react to minuscule changes in light intensity by outputting an amplified antilog signal.
The location of the spot of light playing on the film is mechanically moved. The photomultiplier is also moved in step with the focused light source. These movements and the spin of drum generate an analog representation of the film’s image. To record a color image, three scans are made by passing the light through red, green and blue separation filter. If the image is a print on paper, the light reflects back into the cylinder after traversing the print emulsion. It is then directed to a photomultiplier inside the cylinder.
The analog signal could be used by devices that exposed photo film or photo paper. Some accepted the analog signal, otherwise the signal was converted to a digital signal. These exposing devices were constructed using red, green, and blue laser light sources that were caused to trace a path across film or paper. Various analog and digital storage methods were used. These included giant 10 inch floppy disks and various hard disks not far removed from the ones used today.
I'm going to address one aspect of your question, since most of the rest has already been answered.
I see photoshops offering drum scanning to 1 Giga Pixels and even more. How does it a machine from the past able to extract so much information?
The one gigapixel scans are usually of large format negatives. There's a LOT more information that can be contained in a high quality piece of film that is 8 x 10 inches (20 x 25 cm or 200 x 250 mm) than in a 36 x 24 mm frame of the exact same film in 135 format ("35mm").
The surface area of an 8x10 negative is roughly 50,000 mm².
The surface area of a 135 negative is 864 mm².
One 8x10 negative can contain the same amount of information as 58 frames of 35mm film!
A one gigapixel scan of an 8X10 format image translates into about 17MP per any 36x24 mm area within it. Two gigapixels for an 8x10 format would be about 35MP for 135 format, and so on.
A partial answer:
Those scanners (in begin of 90s) are managed by minicomputer (power of PGP-11) and store digitised images on magnetic tape. For example such scanner produce for 10" photos image with average size 1GB and store the information on magnetic tape like this. The read/write unit is something like this.
About the position of the lamp - it can be also inside when you scan film or plate.