The problem is, nobody is quite sure how PIE was pronounced!
When we talk about PIE phonemes like
/*d/, we don't mean it was actually IPA
[d]. We mean that "there seems to have been a phoneme, which is pronounced
[d] in a lot of descendant languages". But there are also many languages which don't pronounce it
[d]: Germanic, Anatolian (an extremely conservative branch), Armenian, and Tocharian all have
[t] there. So there's a theory that
/*d/ was actually
[t'], an ejective, which makes certain weirdnesses in PIE a whole lot less weird. (Look into the "glottalic theory" for more details; there's some info in this question's answers too.)
Similarly, PIE has two "velar" series, the ones marked with accents (
/*ḱ/) and the ones without (
/*k/). It's clear that there was a difference between these series, because they act differently in the satem languages, but there's very little agreement on what the difference actually was. Personally, I like the theory that
/*ḱ/ was velar (
/*k/ was uvular (
[q]), but you can find a great number of linguists who would vehemently disagree with me. In particular, if you don't accept the glottalic theory, the uvular theory makes the weird consonant system even weirder, requiring sounds that aren't attested in any language in the world.
On the plus side, there are only two (maybe three) phonetic rules reconstructed for PIE that aren't represented in written-out reconstructions:
/*e/ next to
/*e/ next to
And sometimes (depending on reconstruction):
- Consonant clusters assimilate in voicing, and nasals assimilate in place
That change definitely happened at some point, there's solid evidence for that, but it's unclear if it should be called a part of PIE or a later change.
So if you want to turn a PIE word into IPA, the process is straightforward:
- Decide which pronunciation you want for each reconstructed phoneme
- Map each phoneme onto its chosen pronunciation
- Apply the two (or three) sound changes
And you're done! But note that this pronunciation is very definitely not accurate to what the actual early Indo-Europeans would have used: that's the problem with using reconstructions. There's just too much detail that's been lost over the millennia. (In particular, there were certainly more phonetic changes than the three I listed—we just don't have enough evidence to reconstruct others with any certainty.)
Here's my personal guess as to pronunciations:
But it must be emphasized that this is only a guess: an educated guess, sure, but there's just not enough evidence to call it a "theory", and given the lack of evidence, it's not really falsifiable. Other linguists here can probably give you a dozen other reconstructions that differ from mine. That said—while this guess isn't better than any other, it's also not really worse than any other. So depending on your purposes, that might be enough.