Arve: (going to redo the relevant bits of syntax when I get home, but)
Des Conlangery leich vön sattes.
In Standard, that'd be (was going to give both IPA and X-SAMPA because my phone hates Unicode, but nooo, I can't even copy text in an input field without copying the whole fucking field, so
http://conlanger.com/xipa.html )
[%CA~O~_^%nla~NI"rIj "lVi_^M\_0 Y "tr_0_hV?Is]
But you could also get away with Midland, which is:
[%Ca~@~_^%l{~NI%rIj "lVi_^C Y "r_0V_ki_k_^s]
So do whichever's easier. Though prosody is unintuitive enough that it'd probably be better to wait until I'm at home and can do a recording... Essentially, stress and pitch go up, instead of down, in a prosodic block(?) -- i.e. "con", or rather "des con" is lower and less stressed than "lang", which is about equal, though possibly lower, than "leich". (Why doesn't IPA have a symbol for tertiary stress?) But at the same time, "con" and "ry" are longer than "lang" and "leich", because they don't have an unstressed syllable attached to them. (I'm not sure how to handle the problem of the Midland syllable loss in "sattes". I'd use it to reintroduce vowel length, but there's enough going on already, and I've not elided anything else in that position so all the long vowels would be creaky and all the creaky vowels would be long.)
Northern is much easier, though, but I've forgotten how it works, and the grammar would be simplified in ways that I've not worked out yet. (As a side note, I really want to see a conlang that innovates phrases as a result of prosody changing reduction patterns, as happened IMI with "I've not", which I think exists because it's one syllable shorter than "I haven't", as well as avoiding an ugly vowel cluster. Though it appears more in writing for me, so it's probably actually just that it's shorter.)