If your singer is that soft when she sings, be very careful with a compressor on her voice. First, if she is compressed much, then when she stops singing her mic will float back to the higher uncompressed gain stage which will definately increase the chances of feedback on stage. Next, if the band is really loud you may have a bigger issue. Take a look at the meters on the vocal channel. Do they go higher when she sings? Or is the band already so loud that they don't change. If so , then you may have to consider a plexi for the drummer (if he won't change) and different amp placement and/or volumes. This is actually very common when dealing with softer singers in rock bands. Its pretty much physics. If the snare and guitars hit the vocal mic at a higher spl than the vocals, than that mike is as much or more of a "room" mic than a vocal mic. A compressor may actually be exaggerating this problem and will result in less clarity. Making the move to in-ear monitors will definately help. First, no more wedges to create additional noise and feedback loops. Second, players will tend to be a little more relaxed when they have to turn down their amps a little since they can still hear it so well. This will GREATLY increase the speed and ease of which a decent mix is brought up at FOH and add much more clarity and presence to it. Consider though that you will most likely need to add a kick, a snare and at least 1 overhead mic if you move to inears so everyone can hear the drums properly. The mics don't necessarily need to be in the mains though, just the monitors
Asfar as the Behringer comps go, they really aren't a bad deal. In no way would I consider them a good compressor, but in my opinion they do sound much better than they cost. I am going to guess that your PA is not actually that large since you mentioned that the drums aren't amplified. Personally, I would do everything possible to avoid compressing your mains. Regular old compressors are not designed to really be inserted on program material. Its very easy to get out of control and pump things really bad. All it takes is one loud bass note well your singer is singing to launch the compressor. What will happen is the bass note might trigger the comp which in turn will compress (lower the volume) of the whole mix. In turn, the vocal will actually be dropped in volume because of the stray note which is what you really wanted dropped. I would start with compressing the bass, and very cautiously compressing the vocals. I wouldn't really worry about comping the guitars, at least not yet. If the guitars are really needing it, thats more likely a sign if improperly balanced channels (on the guitar amp itself). Also, I wouldn't be too worried about blowing your ears out with the in ear monitors. Having the inears will tend to make performers dynamics a little better. You will know when you have gotten too loud and back of yourself. This is why I very rarely compress a singers voice in their monitor. If they are being really spotty with their volume, then I WANT them to know it. By compressing their voice you may actually be encouraging them to use improper dynamics which increase the problem at FOH.
On a smaller PA, I don't think the manual attack and release settings are really all that necessary. They are nice to have, but not a necessity. Maybe consider one auto com and one multi com if you really need 6 channels. Or better yet, save up a tiny bit more and get a DBX 1046, or maybe one autocom and one DBX 166. There are a lot of options available, but the Behringer comps are a pretty good value
Sorry for the long winded post, hope it helps
