Telescopes this large are a formidable engineering challenge. The huge 8.1 metre glass primary mirror must be lined up precisely with a target that may be billions of light-years away. In order to bring the light to the sharpest possible focus, the whole telescope must hold its shape to an accuracy better than one tenth the wavelength of light at all times! This means that the whole telescope, though the size of a large house, must work to a precision better than a Swiss watch.
The first step takes place during the afternoon prior to the observations. Telescope operators use the weather forecast to estimate what the likely night-time temperatures will be that night. The inside of the dome is air-conditioned down to this likely temperature. If this were not done, and the telescope or dome was warmer than the outside midnight air, this would cause thermal air-currents which would blur the images received.
At the start of the night (and continuously during the night) the main 8.1 metre mirror must have its shape "tuned". The mirror is too thin (20cm) to hold its parabolic shape to the necessary accuracy all by itself - any mirror that would be thick enough to do this would store up too much heat during the day, and would hence produce thermals all night, blurring the images. Instead, the mirror is pointed at a bright star, and the light collected using a wavefront sensor. This calculates how much the mirror is out of shape, then 180 computer controlled actuators push and bend the mirror until it is in perfect shape.
The 342 tonne telescope is then pointed at a target. This isn't easy - the telescope needs to track objects with an accuracy equivalent to the apparent size of a golf ball 300 km away. And as the telescope tilts over, it bends under its own weight: there is no way anyone could build a 342 tonne object that didn't bend at least a little. It also changes shape as the temperature of its steel supports changes. All of this has to be allowed for and corrected for using a computer model.
The mirror is tuned, and it begins to collect light. The mirror is made of 22 tonnes of ultra-low-expansion glass - glass that does not expand or shrink as the temperature changes. Such changes would warp its shape and distort the images. The mirror is coated with a thin layer of silver, with a transparent protective coat on top of it to prevent it tarnishing. This silver layer reflects the light up to a secondary mirror, hanging suspended high above the primary. The mirror is polished so well thata the typical irreularities on the surface are only 15.6 nano-metres is size (ie. about 100 atoms thick). If Australia was as flat as this mirror, the biggest mountain would be less than a millimetre high!
The secondary mirror bounces the light back down through a hole in the primary mirror, to the instruments. Nobody actually looks though modern professional research telescopes - the light is captured by a digital detector hundreds of times more sensitive than the human eye. The Gemini telescopes have a wide range of these digital scientific instruments. Some take pictures, at ultra-violet, optical or infra-red wavelengths. Others break the light up into its component wavelengths (spectroscopy), using prisms or diffraction gratings.
But before the light is sent to the detectors, one final step may be applied: adaptive optics. Even the best telescopes at the world's finest observing sites are still affected by the atmosphere. Thermals in the air over the mountain warp the incoming light waves, blurring the images we can obtain. Because both Gemini telescopes are superbly located, this fuzziness is only half as bad as at some other sites. But it still limits what we can see. The solution is to find out precisely how the atmosphere is distorting the incoming light, so that the telescope system can "un-distort" it.
Gemini does this by firing a laser up throgh the atmosphere. The laser is tuned to a wavelength that causes sodium to glow. Between 80 and 100 km above our heads, in the tenuous upper reaches of our atmosphere, it turns out that sodium atoms are concentraed (they probably got there in a rain of microscopic meteorites). As the laser beam ploughs through this layer, it causes the sodium atoms to glow brightly, creating a temporary "artificial star".
The light from this artificial star is observed by the telescope. This light is so bright that the distortion in it due to the atmosphere can be measured 50 to 100 times per second. These distortion measurements are fed to a small deformable mirror, which bends in exactly an equal and opposite way. The light gathered by the telescope is bounced off this mirror, cancelling out the distortion, and producing images that can be ten times sharper than before! Unfortunately, the distortion changes very fast, so the deformable mirror needs to change shape a hundred times a second: but this is now possible.
Impressive though the telescope is, the greatest challenge is usually the instruments that dangle off the back and record the light. Many of the targets Gemini studies are so faint, that the whole enormous telescope may only collect a few photons of light from them an hour. The energy trickles in so slowly that you'd have to collect this light for over a trillion years to get enough power to run a normal light bulb for one second! Typical instruments use a combination of high precision lenses, gratings and mirrors to dice up the light into its constituent parts and focus it on a low-noise detector: a cryogenically-cooled piece of solid state physics wizardry that actualy records individual photons as they hit.
The telescopes are run from a control room, well away from the actual
dome (people in the dome would radiate far too much heat, causing
thermals and blurring the images). The telescope operator and
astronomer slew the telescope from target to target, using networks of
computers. Perhaps the most complex part of the whole telescope is the
software that drives it, and which records and analyses the data.
Australian Gemini Office, ausgo -@- aao.gov.au

The Australian Gemini Office (AusGO) is operated by the
Anglo-Australian Observatory (