1.Get word, syl, phone times using dumpfeats
2.Get TextGrid with all feats for raw speech
3.Annotate words, syls and phones  with breaks, prom and Intensity peaks and valleys respectively
4.Dump in bracketed format in voicedir/prosdoy_feats as .break, .prom, .Ipeak and .Ivalley
5 add word, syl and phone feats to utts
6. Add appropriate pathnames to mcep.desc

Phrase Breaks:
( R:mcep_link.parent.R:segstate.parent.R:SylStructure.parent.parent.pt_break float)
( R:mcep_link.parent.R:segstate.parent.R:SylStructure.parent.parent.R:Word.n.pt_break float)
( R:mcep_link.parent.R:segstate.parent.R:SylStructure.parent.parent.R:Word.p.pt_break float)

Prominence:
( R:mcep_link.parent.R:segstate.parent.R:SylStructure.parent.pt_prom float)
( R:mcep_link.parent.R:segstate.parent.R:SylStructure.parent.R:Syllable.p.pt_prom float)
( R:mcep_link.parent.R:segstate.parent.R:SylStructure.parent.R:Syllable.n.pt_prom float)

Intensity:
( R:mcep_link.parent.R:segstate.parent.pt_Ivalley float)
( R:mcep_link.parent.R:segstate.parent.n.pt_Ivalley float)
( R:mcep_link.parent.R:segstate.parent.p.pt_Ivalley float)
( R:mcep_link.parent.R:segstate.parent.pt_Ipeak float)
( R:mcep_link.parent.R:segstate.parent.n.pt_Ipeak float)
( R:mcep_link.parent.R:segstate.parent.p.pt_Ipeak float)

7. Cluster
8. Synthesize

./setup /path/to/voicedir
cd /path/to/voicedir
./do_prosody_build