Joe, responding to Mark J. Reed:
> >        ś       (LATIN SMALL LETTER S WITH ACUTE)

> s-acute = [C]

Eh? I've always been under the impression it was [S] (though Hindi may not
be the same as Sanskrit....)
> [r=]

I do believe this is the syllabic /r/, in Skt. at least by convention
pronounced [ri] or [rI]-- "r.s.i" [r(iI)Si] 'sage, wise man' note the
Anglicized "rishi". That must actually be a quite old pronunciation, as Skt.
loans in Javanese and other SEAsian langs. also have it. Whether it was
trilled in Skt. I know not; nor do I know about Hindi.

"h with" is indeed a slight aspiration-- in Skt. only in final
position, where it represent the -s of various endings.  Root raj- > raj-a-s
written rajah. 'king (nom.)  My Skt. teacher (who'd studied in India) said
to pronounce it with a slight echo-vowel offset, ['radZah(a)]. He also said
that the cluster jñ- as in jñana 'wisdom' is pronounced [Nñ...], but when we
asked where that [N] came from, he just replied "Because." A real nice guy;
one day he reduced a foreign student to tears.

Further to Mark's other queries-- t d n s with a are _retroflex_.
Unlike Mark, I don't find Skt/Hindi over-diacritiized at all (certainly less
so than Schmampa....) and on the whole it's quite reasonable and elegant.
And much of it simply sanctified by tradition (like the n-overdot for N-- a
transcription probably devised early on, before anyone had thought of IPA--
indeed, many 19th C. "phonetic" renderings follow this usage).