Acoustical Space is the term used to describe the apparent environment that a sound "appears" in. That is to say, the sound of the room, the reverb, the materials, etc. A vocal bathed in short bright reverb sounds like it was recorded in a bathroom. The sound of the bathroom is the Acoustical Space we're talking about (whether the vocals actually were recorded in a bathroom is, to the end listener, not relevant.)
There are two differing philosophies when it comes to acoustical spaces:
One says that a mix engineer should try to put each element of the recording (instruments, vocals, etc.) into its own acoustical space. For example, having a fair amount of reverb on the snare drum, but very little or none on the vocals. The theory being that since our ears can discriminate between the spaces, it makes a mix less cluttered, and gives definition to the elements. While this may have some truth to it, most recordings use this theory far too aggressively, resulting in a mix that doesn't stick together, and sounds like 5 or 10 recordings layered one on top of the other.
The other philosophy, which I adhere to, is that there should exist a common acoustical space, and that all of the elements should appear within that space. That is, it should sound like a really great recording of the best live performance the band has ever done. I'm not opposed to special effects and theatrical use of effects, but I think they should be just that, effects. Elements can be brought to the forefront through eq, volume, compression, and the manipulation of "room tone."
Room tone is simply the sound of the room while no one is playing (start recording your vocal mic, with no one singing, and you can hear the room tone.) Once the musicians are playing, room tone is the sound of the instrument in the room, to an observer, not the player. A mic in a corner of the room, 15 feet from the drum set is going to capture the way the drums sound in that space, for example.
Of course, if you're relying on real room tone to give you the acoustical space, you may run into problems with synthetic instruments. In this category, you'll find drum machines, keyboards, guitar and bass amp emulators (Like Line 6 Pods, Amp Farm, etc.). One of the biggest flaws I hear on many recordings is the use of Pods. If your drums and vocals are in an acoustical space that sounds like a mid size, semi-live room, and the guitar is using some emulated Marshall stack in an arena, guess what? It doesn't work. It sounds fake and cheap. Similar problems occur with drum machines and keys. So what do you do?
One solution is to "reamp" the drum machine or keyboard through an actual amp, and mic it up. It you want a good guitar tone, play through a good amp, and mic it properly. This is the method I would advocate, generally speaking. Now, if the instrument in question is the bass guitar, you may be able to get away with a Pod or DI. Bass is pretty non-directional, and we don't get a lot of spacial information from it. But a mic'ed bass cab will sound more authentic than a DI bass.
The other solution to the problem is synthetic reverb. Using a plug-in or an outboard unit to add reverb to a sound is tricky. Sure, it'll sound really cool all by itself. But when you mix those reverbed out vocals in with the other dry instruments, it's going to stick out like a sore thumb. So, you need to understand how reverb works, if you're going to pull it off. Two tricks help a lot. One is to set up a stereo reverb on a send, then run a bit of each instrument through it. You can run more of the snare or drum overheads and less of the guitars and bass. But even if you just add a tiny bit, the listener's brain puts it together and says "ok, that sounds right."
The other trick with reverb is to use a pre-delay. Most of the time, when we hear reverb in real life, it is preceded by a space of silence. Like this:
sound>> (space)>> echo>>reverb
The space can be short. (A rule of thumb is this: Sound travels at about 1100 feet per second. We can round this off, and say that sound travels at 1ft per millisecond.) If you're in the center of a room that is 20x20, then a predelay of 20 milliseconds will sound pretty natural. This is because the sound would leave you, travel 10 feet to the wall, then travel 10 feet back to you, 20 feet round trip. You can use this knowledge to do some funky reverb things, too. Split the signal, and send one through a 10 millisecond delay, and the other through a 20 millisecond delay, then send both to the reverb. You'll get the illusion of being closer to one wall, particularly if you use hard panning.
In general, though, I like to use a real room mic or two. These will be mixed in very low, and probably compressed pretty heavily. Compression is generally used to "even out" the performance, making the louder parts closer in volume to the quieter parts. An interesting side effect of this, though, is that the room tone gets louder and more audible as you compress the signal. Before adding reverb to the drums, try compressing the overheads a bit. This opens up the sound, without muddying things. Of course, if the room sounds bad, you'll want to close mic everything (vocals included), and use compression very conservatively.
This technique of compressing a room mic and bringing it into the mix is often said to "Glue" the mix together. It makes it sound more like a real performance, and less like it was crafted together one bit at a time (I've even heard of engineers running the entire mix out to an amp, mic'ing it at a distance, and mixing that signal back into the mix.) Of course, different genres will have different aesthetics in this regard. While room tone is valued in classical and jazz, it is less treasured in metal and pop. And some genres even value the disparateness of isolated acoustical spaces, hip-hop and some dance genres being an example.
For my money, though, nothing sounds as good as a real band playing in a real room. Close mic everything, and set up a room mic or two. Mix and season to taste!