I’ve been working with Chiyu lately, and if you’re an english speaker that has spent time working with a chinese voice, you’ve probably encountered a few oddities in the phonemes, if not in the way that some entered words sound hilariously like nothing you expected. The english interface for the chinese voices uses pinyin for word entry, which looks like standard english but often maps to sounds in surprising ways. Many of the pairings are counterintuitive (a lone ‘e’ is a schwa, ‘c’ sounds like “ts”, etc.).
The link above provides a good reference for pinyin pronunciation. The site is laid out in groupings, mostly based on the vowels but later on the more complex consonants. Each combination has an example sound file so you can hear what is likely to be generated. It’s all free and public, and it requires no account creation, and it saved me quite a bit of frustration.