There's no established notation for ambisyllabic consonants, but in _The Phonology of German_, Richard Wiese suggests an overdot, as in [fal̇ən] (1996:36).

But keep in mind that syllables are theoretical constructs, so you can choose a different model of the syllable to suit your needs (or theoretical biases).  Some notions of the syllable don't allow for ambisyllabic segments at all, while some crucially rely on them.  There are even some extreme models of the syllable that only allow for strict CV structure, with codas and consonant clusters the result of CV syllables containing "null" vowels.

Think about why you even need to talk about syllables in your language at all, and why it would matter phonologically if a consonant is ambisyllabic.  If there is no phonological importance, and this is really just a question of predictable phonetic implementation of an intervocalic singleton consonant, then you don't really need a special notation for that (just as we don't need a special notation to indicate that the [b] in "bleeding" is predictably shorter than the one in "bee").


> On May 7, 2017, at 9:15 PM, The Scribbler <[log in to unmask]> wrote:
> How do you notate a non-geminate consonant in IPA when it's clearly pronounced across the syllable boundary? Especially affricates when they are NOT consonants in hiatus?
> I'm having a rather painful time trying to write up the pronunciations for this language that has geminate consonants phonemically and is having a ball with starting a consonant in the coda of one syllable and finishing in the onset of the next.