you're probably better off just manually cleaning out those "sss", "tttt', "chhhh" sounds honestly. Depending on how the vocals are stacked you may find a great setting on a de-esser for 75% of a song only to find out that the other 25% the singer now sounds like he has a lisp. It can also dull the vocal track and rob its energy (I find). Anyway, try finding where the main body of the offending frequency is with an EQ and cut about 4-6db there with a not too wide Q but more than just a notch. Then roll through to each spot in the track that has it and print just that spot. If you still need more in really heavy handed sounds, reprint or use a little volume automation. At least this way you have much more control over the end result and you'll use less automation during mixing, saving time as well as not having to use an insert spot or whatever for a de-esser