Sound is subjective. Ultimately, your personal impression should be your guide when selecting the right microphone. Still, technical data can give you valuable information about the usability of a particular microphone.
Noise Performance (Self-Noise)
Noise is an important factor for distant miking, e.g. a choir or an orchestra and for recording quiet sources such as nylon string guitars.
For condenser mics noise performance is usually specified as "self-noise." Alternate terms are "equivalent noise level" and "equivalent sound pressure level." The lower the self-noise figure, the less noise the microphone produces. 16 dB-A or less is an excellent figure for a small diaphragm condenser, 20 dB-A is still pretty good. At about 23 dB-A or more you're likely to notice some background hiss, especially when you're miking quiet sources.
Note: The "A" means the figure is weighted according to the sensitivity curve of the human ear. A-weighted figures emphasize the mid frequencies and disregard noise at the extremes of human perception, i.e. very high and very low frequencies. Some manufacturers also give unweighted noise figures (which invariably look worse). Don't confuse weighted and unweighted figures; they're in no way comparable.
Sensitivity means output level. Various microphones will produce different output levels when exposed to the same sound source at the same distance. The usual way to specify sensitivity is in millivolts per pascal (i.e. output voltage per sound pressure). The higher the sensitivity figure, the "louder" the microphone. A typical small diaphragm condenser mic is about 5-10 mV/Pa. The higher the figure, the less gain is required from your microphone preamp.
Unfortunately, there are several other ways to specify sensitivity, the most common being in relation to a hypothetical microphone. Which apparently has an enormously hot output level, given the fact that all known real world microphones produce negative decibel figures in comparison, usually between -55 and -35 dB. Here, too, higher figures mean higher sensitivity. But don't let the negative numbers confuse you: -35 dB is much higher and therefore "hotter" than -55 dB!
Some people think that exceeding the maximum sound pressure level specified by the manufacturer means your microphone is in jeopardy. Don't worry! Your microphone won't be destroyed, it just means that sound at this level will no longer be free of noticeable distortion artefacts. Max SPL is usually specified as the sound pressure level ("volume") at which THD (total harmonic distortion) exceeds 0.5 %. Some manufacturers use 1% THD as the threshold figure, which results in - seemingly! - more impressive max SPL figures. In home applications and even most professional studio applications, maximum SPL is nothing to worry about. Usually your neighbours will greet you with a rifle, long before your Marshall stack exceeds your mic's maximum SPL. So you better worry about your neighbours ;-)
Some sources can produce very high SPL figures in close proximity, e.g. brass instruments and drums. A few sources are much louder than you'd think because they contain very little energy in the bass department (where you'd physically feel it). A tambourine, for instance, can easily be louder than an average microphone's maximum SPL. 130 dB-SPL is a very good maximum SPL figure, 140 dB-SPL and more is excellent. Oftentimes you can add an additional 10 or 20 dB by applying a pad which reduces the level right behind the capsule, so the internal amplifier electronics won't overload. However, don't use a pad unless necessary. A pad often leads to a slight increase in noise and may also compromise the sound quality somewhat.