Any advice for someone who is new to mastering?

[-ZERO-]

I have been trying to finish the track above so I can finally move on to other projects and I'm looking for advice on mastering.
So far I have been following a couple of YouTube tutorials and I don't understand what I'm doing, and I feel like my lack of understanding is affecting the quality of the track.
I have been using AICA's English voicebank for the vocals, and she is very high-pitched and tends to over-emphasise s and z sounds, and I can't tell if that is because I messed up something during the tuning process (like maybe I made my track an octave too high and that's why the voice sounds so squeaky?), or this is something I can fix during the mastering process, as from one of the tutorials I saw, using compression can help clean up some of those harsher sounds, but it doesn't seem to be doing that despite me having applied it to the vocals.

What bugs me the most is how my own work stands in contrast to AICA's release video in which she sounds very smooth. I understand I probably shouldn't compare my work to the work of a person who has been doing this for years, but I'd be a lot less bothered if I understood specifically why my track sounds off.

Also on a final note, I saw someone a while back say to put the breath sounds during the mixing process rather than while tuning (as putting the breaths through utau will distort them due to the resampler), but I can't find the .wav files for the breaths. I know they exist, as the readme has them listed and if you type them up in openutau they show up, but I can't find the .wav files for the breaths regardless

 

[Kiyoteru]

The term you're looking for is probably "mixing", not "mastering". Mastering would be the process of preparing audio for final distribution, like having the correct file formats to upload into a distributor, or preparing the arrangement of album tracks for pressing onto vinyl records.

For loud consonants you could use a de-esser plugin, which is basically a multiband compressor that only compresses high frequencies. Set the threshold and ratio so that normal consonants are unaffected and the loud hissy consonants get compressed.