So you've successfully notched out the "nasties" from the vocal. Now, as far as the harmonic balance of the vocal sound, you're feeling pretty good. Now, let's listen to the dynamics. Are certain syllables or words popping out? Are the front ends (attacks) of words or lines louder than the rest? Is the volume and presence of the vocal consistent throughout each section? If there are problems in these areas, your vocal might benefit from some compression. Keep in mind, if the song calls for a very natural vocal sound --say, if it's a very open-sounding acoustic track, or a soft ballad-- then we might choose to avoid compression entirely and simply control the dynamics by riding or automating the fader. Compression almost invariably changes the sound of a vocal, making it smaller and tighter... (On a good day, 'focused'!) Sometimes, if the vocal's already sounding good, this does more harm than good. Listen to the song, and listen to the vocal carefully. Ask yourself, what does it need? Compression serves many purposes, but in vocal processing, usually we ask a compressor to even out the volume of words or phrases. That purpose informs how we'll set the compressor. Specifically, if we want the compressor to react to a word that's too loud, we'll set the attack relatively fast so the compressor can react quickly to the word, and set the release quick enough to get out of the way. In this scenario, we'd set the compression threshold to be just above the peak (or RMS) level of the vocals that are at the correct level. Then, when the loud word comes along, the compressor grabs it quickly, reduces it quickly with a medium to high ratio (say, 6:1), and then releases quickly as the vocal settles down. More commonly, we want the compressor to not simply react to specific words, but to keep the performance present and clearly audible in a busy mix. This is how we achieve that coveted "in your face" vocal sound. To accomplish this, we set the threshold lower, so the compressor is working throughout the entire section. Attack will still be relatively fast (do experiment to see what sounds good!), as will release, but these settings will depend on the compression curves of the device or plugin you're using, and how much compression you're applying. Common ratios for this application are 4:1 to 8:1, and you might be applying 3-12dB of gain reduction. Adjust your makeup gain accordingly. Things to listen for: If your attack is too slow, the beginnings of verses or phrases may pop out, as the relaxed compressor "grabs" at the front end of the signal. This can also be a sign that you need to lower your threshold more. A release that is too fast will bring out the ambiance of the room or booth the singer was in, and potentially make the vocal spitty or ess-heavy. A release that is too slow will make the vocal sound somewhat farther back in the mix, reducing the presence that we're using the compressor to achieve. Also, if the threshold is too low, and we're applying too much compression, the effect can be that of abrasiveness or fatigue, as the push and pull of the words and spaces gets too flattened out by the compression and becomes unnatural. Tinker with a compressor on a vocal until you can hear these effects. Listen closely. If you can't hear it... Keep at it! It takes time! Every compressors will react to a given source differently, depending on the compression curves and attack/release characteristics. (If these terms are alien to you, don't worry! I'll do a chapter specifically on these nuances of compressors. For the time being, just embrace that different compressors can behave a bit differently.) When you're setting a compressor, once you get some serviceable settings on the compressor, you'll want to "feel out" where the sweet spot is by listening carefully to the source while adjusting the parameters (I focus on threshold and release). Ideally, you're listening for where the action of the compressor is keeping the source controlled, but also making it more full and exciting. Good hardware compressors will also impart a tone or color to the sound depending on gain staging (i.e. how hard you're driving the compressor's input and output stages). Even plugin compressors will exhibit a sweet spot in the compression behavior if you listen and adjust the parameters carefully! How does the compressed vocal sound in the mix? Is it "up front" and present? Is the sound and performance "focused"? If so, you're on the right track! You may discover at this stage that the vocal needs additional EQ. Perhaps the compression has brought out a nasty frequency that you didn't realize was there before, or perhaps the overall tone now seems too dark. You have the option of adjusting the pre-compression EQ to compensate, or adding another EQ post-compressor (which I do frequently). "2 EQs on a vocal?!? That's insanity!!!" you say. Not really. Remember, our first EQ was really just cleanup. The tracking engineer should have done this for us, but we won't hold it against him. Besides, now that the sound is more focused and present (thanks to the compression), we're better able to adjust the balance of the vocal tone to get it to sit in the mix. This may involve boosting some midrange (say, in the 1-3kHz range) to add some presence and bite, or shelving down some low end (perhaps 300Hz and below) to get the vocal to sit with the arrangement. You may hear a new "nasty" that you want to notch out. Perhaps the vocal needs some top-end brightness that a bell or shelf at 6-8kHz can bring out. Again, you will have to listen, listen, listen to the vocal track, the rest of the mix, and perhaps some reference mixes to discern what is needed for a particular track.