Once we've decided on a place for the observatory and decided on whether we want it to be a wide-field telescope or a deep-field telescope, we will have to decide on the array of instruments to be setup. Again, depending on what our intentions are with the telescopes, the objects that we would like to observe, we will have to decide on a relevant telescope.
For example, if we intend to observe transient events such as asteroids, comets and supernovae, we will need a wide-field telescope. Why, you ask? Well, a transient event, by definition, is an event that wasn't observed at the same place at an earlier time. As such, supernovae are stars that blow up and asteroids move faster than any of the stars do, from our point of view. Therefore, if we observe the same patch of sky for a very long period of time, we are bound to run into interesting transient events. And since there isn't any special direction or patch in the sky, wide-field telescopes are designed such that they can map the whole sky once a week. Again, note that when i mean whole sky, I mean the portion of the sky visible to a ground based telescope. Similarly, if we intend to study astronomical objects to a good angular resolution, we deploy a deep-field telescope with a camera (CCD) containing a large number of pixels, with small width.
Going off track, in the 50's and 60's, before CCDs were developed and widely deployed for astronomical purposes, photographic plates were the choice of an astronomer. A photographic plate would be placed at the focal place to capture the sky and would then be studied to understand the objects in the field of view. Interestingly, some of them are being digitized by the Smithsonian Astrophysical Observatory. This could help in discovering new transient or variable objects. But photometric plates are highly restrictive in the sense that it's a qualitative study of the objects and not a quantitative one. Relative brightness can be inferred by using a reference star in the frame and angular distribution but information about other properties such as temperature, composition cannot be inferred from the plates. Inorder to overcome this obstacle, astronomers used filters. Filters would be placed infront of the photographic plates, filters that restrict the light falling on the plates. There are numerous astronomical filters, a brief overview of which can be found here. The filters restrict the spectrum of light that falls on, and hence absorbed by, the plate. Recall that a stars can be considered as blackbodies and the spectrum of a star can be studied, albeit qualitatively, using such filters. For example, in the case of two stars A & B observed using a U filter (365 nm), if one A is brighter than B, we can infer that the temperature of A is greater than that of B. This inference comes from the understanding that with increasing temperature, an increasing amount of the star's radiated energy falls in the U band. While not obvious, such qualitative analysis has in fact given incredible results, the most prominent of which is the HR diagram. The HR diagram depicts the types of stars and it can be used to understand the evolution of stars and to infer the distance to star clusters or other galaxies. Photometric redshifts is another interesting use of studying astronomical objects with filters. The replacement of photographic plates by CCDs just made field more quantitative.