An addendum to eluyten and wukong888's remarks on the Wade-Giles romanization system for Chinese:

It looks confusing, but there are actually sound reasons why Sir Thomas Wade chose to write p and p' for what Pinyin writes as b and p. When Protestant missionaries were first exploring the sounds of Chinese, they were very conscious of phonetics and the sounds of Sanskrit, which had been (wrongly) promoted as a kind of "original pre-Babel language" at one time. They were well aware that medieval Chinese, as well as Shanghai dialect, had three classes of stop consonants: voiced, voiceless unaspirated, and voiceless aspirated. The sounds written p and p' by Wade are voiceless unaspirated, and voiceless aspirated. Beijing has no true voiced stop initials, even though the unaspirated p may sound that way to the phonetically untrained ear. But medieval Chinese does have real voiced stops and so does Shanghai. The letter b was reserved to write the true voiced stop. That is in accordance with the principles of phonetic transcription that persist even today in the IPA.

The same is true of t and t' (d is voiced), k and k' (g is voiced), ts and ts' (dz is voiced) and so forth.

Few people who use romanization today are aware of the diversity of Chinese or its phonetic history. So they mispronounce Wade-Giles, or anyway they prefer Pinyin, which is now widely used and which is tailored neatly to Beijing phonology. But Wade-Giles is phonetically more accurate, just as it is designed to be. But you have to know the rules.

Personally, I prefer Gwoyeu Romatzyh romanization, myself. I find the arbitrary use of zh, q, and x in Pinyin rather ugly, and I don't like the way i is used as a dummy vowel. But I use it anyway.