> The E-ELT will be the biggest optical and infrared observatory ever constructed, with a primary mirror just short of 40m across.
I’m slowly beginning to realise that the requirements for a large visible light telescope differ from those of a large infrared telescope, despite most visible light telescopes also operating in infra-red. Consider.
1) Adaptive optics works well enough for IR but so far does not work well enough for visible light. That’s because of the resolution needed for adaptive optics in visible light would need to be at least twice as fine as in IR to get to the diffraction limit.
2) There are far more astronomical sources in IR than in visible light.
3) The brightness of IR sources is much less than in V. That means that for IR what is needed is a single extremely large telescope to act as a light bucket. On the other hand, what is needed in V is a number of dishes. We’ve seen this with Keck I and II, with the LBT, and most successfully of all with the VLTI in Chile with its four telescopes.
4) Wide field imaging needs some sort of lens (eg. Schmidt or Gascoigne). This can be done in IR, but needs a speciality glass like fused silica or (better) CaF2 when used for both visible and IR light.
The VLTI has a baseline between main telescopes of up to 130 metres by 60 metres. This is a lot bigger than 40 metres across. 130*60/40^2 = 4.9. So we don’t expect the ELT to have the same resolution as the VLTI for visible light sources. But as a light bucket for IR observations it ought to outperform the VLTI by a factor of (40/16.2)^2 = 6.
To summarise the summary, the ELT should have up to 6 times better infrared performance than the VLTI, but up to 4.9 times worse visible light performance.