LAST month, we got our best ever view of planets orbiting nearby stars. Though this is a great achievement, the planets are much bigger than Jupiter and are in orbits that range from 24 to 119 astronomical units (AU), where one AU equals the distance between Earth and the sun.

The dream is to be able to see planets as small and as close to their host star as Earth is to the sun. That requires a telescope that can see objects nearly 3000 times smaller than those seen last month, and one that is not blinded by the host star's light - feats that are not possible with even the largest telescope today, the 10.4-metre Gran Telescopio Canarias in Spain's Canary Islands. But in less than a decade, a trio of gigantic telescopes will be able to carry off the task with ease.

The 24.5-metre Giant Magellan Telescope (GMT), the accurately named Thirty Meter Telescope (TMT) and the 42-metre European Extremely Large Telescope (E-ELT) will each collect enough light from these extrasolar planets to allow astronomers to study the composition of their atmospheres using spectroscopy. "Are there Earth-like planets in the habitable zones of nearby stars? That is one of the big questions we'd like to answer," says Markus Kissler-Patig, who works on the E-ELT at the European Southern Observatory in Garching, Germany.

The telescopes might also be able to study supermassive black holes at the centre of galaxies, by mapping in detail the velocity of stars in their vicinity. Today's telescopes can only carry out such measurements on the black hole at the heart of the Milky Way. "With the TMT, there should be quite a few black holes at the centre of nearby galaxies that we ought to be able to study," says TMT scientist Jerry Nelson of the University of California, Santa Cruz. The telescopes might even be able to directly measure the expansion of the universe (see "Beacon at the edge of time")

Although the three telescope teams share the same goals, they are taking radically different approaches to achieving them, either in the design of their primary mirror or in the technology used to remove the blurring or "twinkling" caused by atmospheric turbulence.

The main challenge in building any telescope is its biggest mirror: the primary, the size of which determines the telescope's resolution. The primary collects starlight and focuses the wide beam of light it receives towards a smaller secondary mirror. This then focuses the light further and sends it on to a tertiary mirror, which directs the beam towards one of the telescope's detectors.

The biggest primary mirror that can be cast from a single block of glass today is 8.4 metres across, partly because a mirror any wider would be too heavy and difficult to manoeuvre. Also the greater depth of a larger mirror makes it almost impossible to ensure that it is the same temperature throughout. That is a problem because if different parts of the mirror are at different temperatures the image quality will degrade. The only way, then, to build a bigger telescope is to have a primary consisting of a mosaic of smaller mirrors.

The GMT will have seven very large mirrors (see "Future telescopes"), each made of a Pyrex-like material with a honeycomb structure, which reduces weight while providing strength. Air at a controlled temperature will be pumped into the honeycomb, bringing the entire mirror into thermal equilibrium in 20 minutes. That's not bad, considering that the 100-inch telescope on Mount Wilson in California, which saw first light in 1917, took a whole night for its 33-centimetre-thick primary mirror to reach a uniform temperature.

The TMT's and E-ELT's primaries will have much smaller segments than the GMT, inspired by the success of the twin 10-metre Keck telescopes on Mauna Kea, Hawaii. Going with smaller segments has its advantages, not least that each piece is thinner and easier to manufacture. The downside is that it's much harder to keep all the segments in perfect alignment as the telescope moves. So-called edge sensors are needed to keep track of any displacement between the segments, while large numbers of pistons, or actuators, will push or pull each segment to keep the primary mirror's curvature precise to within a few nanometres.

The other significant technology that these telescopes will be exploiting is adaptive optics (AO). The various layers of the atmosphere, which are at different temperatures or move at different speeds, can distort the light reaching the telescope.

Today's telescopes have add-on AO systems that monitor either a guide star or an artificial star created by firing a laser into the upper atmosphere. Software compares the image of the star with the expected image to work out the atmospheric aberrations, which are then corrected for in real time using a deformable mirror. This mirror comes after the tertiary and is thin, flexible and usually a few tens of centimetres across. It changes shape between 50 and 100 times a second to compensate for the effects of the atmosphere.

However, such add-on systems mean there is an extra mirror in the mix. That's not ideal, because every mirror absorbs some light, reducing the amount that gets to the instruments, and also adds thermal noise of its own, affecting observations in the infrared.

The GMT's solution to this is to turn the secondary into the deformable mirror. "This means you haven't lost any [light], because you would have to bounce light off the [secondary] mirror anyway," says Pat McCarthy of the Observatories of the Carnegie Institution of Washington in Pasadena, California, a member of the GMT consortium. However, secondary mirrors are large, so making them deformable presents a considerable challenge. Mindful of this, the E-ELT's designers are sticking with the smaller quaternary mirror approach. The TMT will have a fourth mirror too, but to reduce the thermal noise the light will leave the telescope and enter a separate AO system chilled to -30 °C.

Despite the challenges, each telescope team is aiming for first light by 2017 and the chance to open a new era for astronomy and cosmology. "We are going to depths of the universe and to a spatial resolution that we have never achieved before," says Kissler-Patig. "You are likely to discover things that you absolutely hadn't expected."