A classical violin and piano piece is not the sort of source material where I would ever use compression. You would usually want to capture the whole dynamic range achieved by the performers, assuming they are sufficiently experienced to balance their own sound as a part of their performance. You might just consider an analog limiter to guard against hitting full-scale in the recording process, but that is not much more than an excuse for sloppy engineering.

After performance and instrument quality, the big things here are room acoustics, microphone positioning and microphone choice, probably in that order. Only after that comes selection of pre-amp, converters and other things such as what small EQ adjustments you make in post processing the recording. Believe me, it does not sound good to have a compressor flattening the performance dynamics and pumping the room echoes up and down at the ends of phrases. Similarly, you should try to avoid "screechy" sounds by exercising your skills as a recording engineer in correct choice and placement of the violin mic.

You can make good recordings of piano and violin using a simple stereo pair in the right acoustic space, but it make no sense to capture a great performance and then perform post processing in a poor monitoring environment. My suspicion is that your perceived need for compression stems from shortcomings in your monitoring conditions rather than directly from deficiencies in the original recording.

In an ideal world, I agree with you. However, what I have to work with right now is far from ideal. The room acoustics in particular stink (13ft x 10ft rectangular room, sheetrock walls, 4 large windows, no treatments). And this is just a high-school student recording a particularly difficult piece she's been working on, not a pro by any stretch, and doesn't have much experience recording or playing alone with an accompanist.

Engage the analyzer check box on the visual EQ interface, for visual representation of the waveform in real time.