I have some perpetual problems. Number one is finding a high enough level that will not clip, ...
You record with 24 Bit. No need to drive the inputs of the interface hot. Stay well below zero and enjoy stress-free recording with good results.
... yet will allow the singer to hear his or her voice with the music without having to turn the music way down. Is a headphone amp the answer?
A good (!) headphone amp is always an advantage because the better the sound is, the better is the performance of the singer. But the question is not only to hear the own signal but rather to hear it exactly as loud as necessary, embedded in the playback and other signals which may come from other musicians.
Two rules:
Recording signal loud, playback quiet:
The singer gets very sensitive, listens and adopts to his own articulations and the exaggerated texture of his voice in the phones. Good for intimate songs and many ballades. The singer may sing a little flat (too low).
Recording signal quiet, playback loud:
The singer fights against the playback. He will try to get louder for two reasons. First, he wants to hear the own voice better. Second, he thinks that he must cut through the playback. Good for harder songs and more excitement. The singer may sing a little sharp (too high).
You already see the direction we are going? A headphone amplifier is a tool, but the balance of loudness is the clue. All signals should be in perfect balance for the intended result. You don't need one amp, you need a mix. This is called the monitor mix. Some people call it "the most important mix of the day".
If so, how do I get separate headphone mixes from Logic out into separate outputs on the headphone amp?
Does your audio interface have a software mixer? I guess so. Then mix the incoming microphone to the playback of Logic. Send the combined signal out via the main output or through another output pair if you want to keep the mains for general work and the other pair for monitoring (which is the preferred method). And, if necessary, make a submix in Logic with exactly the sound you want for recording. This is not always the sound of your raw mix.
There are a lot of other possibilities but I think the above will do for a while.
I am also trying to understand comp tracks.
...
I never feel as though I know where I am when I come back to a comp track with a bunch of takes on it. Does anyone have any suggestions on how to gain more clarity and effectiveness with this tool?
Sounds as if you usually record yourself. This is what I do also and I use comp tracks in a very simple way:
I record a long part or a whole track several times. Then I go back to the machine and use the first take because all others are crap
No, not always. But although I generally like the possibility of total recall, I am more pragmatic when it comes to recording. I search for rhythmical, tonal or expressional failures, from the rest I select what I like best. And then - flatten and merge. No mercy, no whining, no hesitation. It's music, not science, the take is on tape, point. Of course I run into situations where I cannot decide which part to take. But honestly, these are never situations where I have to choose between the best and the very best. Rather the opposite is true, if I have to work half an hour with the comp of a single take, then something is wrong. I may not be in good shape or the emotions do not fit to the song. Better to throw the takes away and try again.